GB2580312A

GB2580312A - Structural assembler

Info

Publication number: GB2580312A
Application number: GB1821200.1A
Authority: GB
Inventors: Tedbury Ivo; Vaughan Felix
Original assignee: Semblr Tech Ltd
Current assignee: Semblr Technologies Ltd
Priority date: 2018-12-24
Filing date: 2018-12-24
Publication date: 2020-07-22
Also published as: WO2020136563A3; WO2020136563A2; GB201821200D0; GB201914482D0

Abstract

The automated assembly system comprises at least one robot 1000 configured to assemble a structure larger than itself. The structure comprises a plurality of structural units, such as building bricks or blocks, which are connected together. The structural units may be connected with connecting materials, such as glue or mortar. The robot is configured to manoeuvre around a structure by climbing and/or traversing the structure itself, as well as assemble it. The robot is configured such that it may carry at least one type of building material when manoeuvring around the structure. Later embodiments relate to a container for transporting a collection of structural units. A method for organising at least one robot to assemble a structure out of the structural units. A method of organising a robot in an assembly system. A staircase for use with an assembly system, and a bypass system for use with an assembly system comprising at least one robot.

Description

Structural Assembler

Technical Field

The present invention relates to systems and methods for assembly of a structure. More specifically, the invention relates to systems and methods for automatic assembly of a structure comprising a plurality of structural units. In particular, the invention relates to systems and methods for the automatic assembly of a structure comprising a plurality of bricks.

Background

Structural units, such as bricks or blocks, may be connected together in order to assemble a structure comprising hundreds, thousands, or even millions of structural units. For example, a backyard pizza oven may comprise around 250 bricks; a house may comprise around 10,000 bricks; and a very large warehouse may comprise around 27 million bricks.

The process of assembling a structure using a plurality of structural units is, and has been for millennia, a laborious, time-consuming, and often dangerous manual activity for human workers.

To aid human workers in assembling structures, machines are sometimes used. For instance, to lift structural units from a supply point and place them in position on a structure.

However, conventional machines such as industrial arms, SCARA and Cartesian robots have a finite working zone, determined from a 'global' origin of the robot (e.g. the calibration point). This fundamental infrastructural constraint limits the form and size of the buildable product. These machines can be mounted on a carriage which traverses along a gantry system, but this still results in a single build area which is equal to or larger than in size to the target structure. Another adaptation is to mount the machine on a mobile platform equipped with caterpillar tracks or wheels, but this limits the working zone to traversable floor surface, distinct from the structure being assembled.

The inventors have developed systems and methods which overcome the problems with known assembly systems.

Summary

The invention is defined in the claims.

In a first aspect, the invention relates to systems and methods for assembling a structure out of a plurality of structural units, wherein the system comprises at least one robot.

In a second aspect, the invention relates to a robot for assembling a structure out of a plurality of structural units.

In a third aspect, the invention relates to hardware or additional hardware for use in an assembly system, wherein the assembly system comprises at least one robot.

In a fourth aspect, the invention relates to a system and method for controlling a robot as the robot assembles a structure out of a plurality of structural units.

In a fifth aspect, the invention relates to a system and method for orchestrating a plurality of robots to assemble a structure out of a plurality of structural units.

The robot of any aspect may be configured to manoeuvre around the structure by climbing or traversing the structure.

The robot of any aspect may comprise a robot body; a first foot coupled to the robot body and a second foot coupled to the robot body, wherein the first and second feet are configured to move relative to the robot body and each other; and an effector coupled to the robot body. The first and second feet may be on opposite sides of the robot body.

The robot of any aspect may include a connecting material storage and delivery system, a structural unit positioning system, a structural unit position correcting system, a robot balance system, a collision avoidance system, and/or a robot positioning system.

The robot of any aspect may be configured to carry structural units, such as masonry bricks.

The robot of any aspect may be configured to obey valve constraints which determine in which directions the robot may travel on a path.

The robot of any aspect may be configured to perform checks at checkpoint positions on the structure.

The additional hardware may be a bypass system for placement in an opening in the structure, wherein bypass structure provides a continuous path from one side of the opening to the other.

The additional hardware may be a path for placement between points in an assembly system.

The additional hardware may be a staircase system, where the staircase system provides a plurality of steps between a plurality of access levels. The access levels may provide access to a structure, or access to a pallet of structural units. The plurality of access levels may be at heights corresponding to, say, every second, third, or fourth step. In a preferred embodiment, the plurality of access levels are at heights corresponding to every third step.

The structural units may be masonry bricks.

The invention may comprise one, some, any or all of the features described herein, and if more than one such feature, those features in any appropriate combination.

Brief Description of Drawings

Figure 1 shows a close-up diagram of an exemplary robot.

Figure 2 shows twelve exemplary diagrams illustrating the intermediate configurations of the robot joints during a step-up action onto a brick.

Figure 3 shows twelve diagrams illustrating exemplary intermediate configurations of the robot joints during a step-up action onto a brick.

Figure 4 shows a perspective view of an example mortar storage and delivery system.

Figure 5 shows a perspective view of an example nozzle fitting attached to the end of the mortar container cylinder.

Figure 6 shows a perspective view of an exemplary mortar storage unit, robot unit and custom dispenser apparatus which interfaces between the two.

Figure 7 shows perspective views of examples of a robot refilling itself with mortar from a storage unit.

Figure 8 shows a perspective view of the exposed workings of an exemplary robot gripper 82040, shown clasping a typical brick.

Figure 9 shows a perspective view of the underside of the exemplary robot gripper, shown without a brick in place.

Figure 10 is a diagram of the exemplary robot configuration just after it has spread the mortar for a brick to be laid.

Figure 11 is a diagram of the exemplary robot configuration just before placing a brick on a brick wall.

Figure 12 is a diagram of the exemplary robot configuration as it places the brick in the mortar.

Figure 13 shows twelve diagrams illustrating the intermediate configurations of the exemplary robot during the brick placement action.

Figure 14 shows a perspective view of an exemplary pallet of bricks which have been specially prepared to allow robot access.

Figure 15 shows the above exemplary pallet after it has been packed.

Figure 16 shows an exemplary large container for the transportation and storage of a large volume of bricks.

Figure 17 shows a perspective view of an exemplary storage area and integrated staircase for bricks on site.

Figure 18 illustrates at a high level an exemplary software process taking a BIM model design of a structure all the way through to the actions of the robot.

Figure 19 shows an exemplary flow diagram for the software the computes the robot's location based on a range of sensory information.

Figure 20 is a diagram illustrating an exemplary high level control flow for the robot when placing a brick.

Figure 21 is a diagram showing example situations where valves and checkpoints are used to control the flow of robots.

Figure 22 shows a perspective view of an exemplary robot vertical transportation device or lift.

Figure 23 shows a side elevation view of an example robot vertical transportation device.

Figure 24 shows an exemplary sequence of perspective views of an example robot vertical transportation device.

Figure 25 shows an exemplary staircase that enables the robot to access different heights of the brick structure.

Figure 26 shows a side-on view of the exemplary staircase design.

Figure 27 illustrates the design of the staircase shown in Figure 25 positioned in front of a brick wall Figure 28 shows a perspective view of an exemplary bypass structure on its own.

Figure 29 shows a side on view of an exemplary bypass structure.

Figure 30 shows an exemplary perspective view of a design for a bypass structure, here shown enabling the robot to bypass a window gap in the brick wall structure shown behind.

Figure 31 contains a diagram showing a plan view of an exemplary construction site during the build process.

Figure 32 shows an example of a unit designed to allow the robots to charge their batteries autonomously.

Figure 33 contains three exemplary diagrams of the possible network connections between the robots and a central controller.

Detailed description

Automated Assembly System Overview The automated assembly system comprises at least one robot configured to assemble a structure larger than itself. The structure comprises a plurality of structural units, such as bricks or blocks, which are connected together. The structural units may be connected with connecting materials, such as glue or mortar. In a preferred embodiment, the structural units are bricks, and the connecting material is mortar.

The robot is configured to manoeuvre around a structure by climbing and/or traversing the structure itself, as well as assemble it. The robot is configured such that it may carry at least one type of building material when manoeuvring around a structure. The robot is configured to manoeuvre to a position on the structure whilst carrying building material, add at least part of the building material to the structure, and then manoeuvre away from that position on the structure. The robot may be configured to carry both a structural unit and a connecting material. The robot may be configured to add a structural unit and the connecting material required to connect that structural unit to the structure before manoeuvring away from a position on the structure.

The robot is controlled by an individual robot control program. The robot comprises at least one on-board sensor in communication with the robot control program.

The robot control program may be configured to use the metrology of the structural units as discrete objects for orientation, location, and/or locomotion.

The robot may be configured to define its origin locally on the structure. At least one onboard sensor may sense the robots real world environment. The robot may then simultaneously construct a map of its environment and keep track of its location within it.

The robot is able to self-correct its local positioning by comparing the real world environment it senses with other sources of information, for example a model of the structure, imaging of the structure, or other sources.

The robot may also include at least one on-board sensor to sense the robot's position in global space. Using the knowledge of where it is in global space, the robot may self-correct its local positioning.

The robots may be connected to an energy source. Alternatively, the robots may comprise an energy source. For example, a robot may include a battery. The battery may be charged at a robot charging station.

The automated assembly system may include additional hardware on which robots may manoeuvre. For example, the automated assembly system may include additional hardware which robots use to climb up and/or down the structure, additional hardware which robots use to manoeuvre from one part of the structure to another, and/or additional hardware which robot use to traverse and/or climb up/or down areas off the structure. The automated assembly system may include also include additional hardware configured to allow robots to bypass openings in the structure, such as gaps for windows and doors.

The robots may be configured to use many possible types of building material including, structural units, materials for connecting the structural units, and/or any other material used when assembling a structure. For example, bricks, mortar, and/or wall ties.

The robots may be configured to receive building material at a supply point. A material supply point may supply a single type of building material, or multiple types of building material. For example, a structural unit supply point supplies a robot with structural units, such as bricks. A connecting material supply point supplies a robot with materials for connecting the structural units, such as mortar. A wall tie supply point supplies a robot with wall ties. A combined supply point supplies a robot with a combination of building materials.

The robots may be configured to walk to the supply point to receive building materials. For example, the robot may be configured to receive building material at a supply point, manoeuvre to a location on the structure whilst carrying the building material, add at least part of the building material to the structure, and then manoeuvre away from that location on the structure. The robot may then manoeuvre to a supply point to get more building material and repeat.

The robot may be configured to always walk/climb to the same supply point, or it may be configured to visit different supply points. For example, a robot may be configured to visit a first supply point to receive a first type of building material, and a second supply point to receive a second type of building material, wherein the first type of building material is different to the second type of building material. In this way, supply points may be located and/or sized relative to how often a robot will need to resupply a given building material type. Further, the supply point may be optimised for supplying the given building material type.

Consequently, the speed at which a structure is assembled is increased. In another example, a robot may be configured to visit a first supply point before visiting a first location on a structure, and a second supply point after visiting the first location on the structure, wherein the first and second supply points are different. In this way, the flow of robots in the automated assembly system may be improved.

The robots of the present invention may be configured to use standard brick masonry as the structural units for assembling structures. There are several reasons for this: prevalence despite being an ancient construction technique, brick is still widely used today as a cladding and structural material; demand -with increasing labour costs and a shortage of skilled workers, there is a desire to reduce the number of skilled manual workers; and integration into industry -the technology works with the well-established materials, designs and processes of used in conventional bricklaying, meaning that the invention works within existing building regulations. In one example, a robot is configured to manoeuvre, pickup, carry, and place bricks, carry and dispense mortar, apply a wall tie, and clean the excess mortar to leave a neat pointing style.

The robots of the present invention may be distributed, working in teams and making use of designed collective behaviour principles to coordinate their movements. Adding more robots, more material supply points, and more efficient robot flows/coordination increases the speed of assembly. More and more robots can be added to the team until a saturation point is reached where there is no more space in the automated assembly system.

The automated assembly system of the present invention removes the height constraint of ground-based robots, allows for greater local accuracy, and more economical power consumption than known assembly systems. The advantage over traditional off-structure robots is that the robots are far smaller and cheaper than would be required to assemble a structure of the same size.

When compared with conventional method for assembling structures, the automated assembly system of the present invention may increase the speed at which a structure may be assembled, the complexity of the structures which may be assembled, and the accuracy with which structures may be assembled, as well as reduction or removal of labour costs.

Further, it has the additional benefit of improved safety and well-being of construction workers, removing them from hazardous environments and alleviating them from strenuous and/or repetitive physical labour.

From an economic point of view, this offers a way around a recurrent barrier to wider industry adoption of robotics: up until now, construction robots have typically been large and expensive, forming a high-risk capital investment for construction companies, developers and builders. The robots of the present invention, are smaller and therefore potentially cheaper, and this will lower this barrier to wider adoption in the construction industry.

Robot Overview The robot is configured to perform set of actions for the purposes of automatically assembling a structure. For example, the robot may be configured to pick up bricks and mortar from a material supply point, manoeuvre around the structure to a position on the structure, apply the mortar, place the brick, and then climb back to a material supply point or a charging point. The robot's actions are implemented by its individual robot control program. The robots actions are determined by analysing a model of the desired construction, such as a 3 dimensional (3D) model of the structure. In one example, the robot actions are determined by analysing a Building Information Model of the structure. The robot may analyse the model of the desired construction. The robot may determine a map of its local environment and compare the map with a model of the desired construction. The robot may use this comparison to determine its location and/or determine that alterations should be made to the structure. For example, the robot may determine that a brick should be moved, or that the amount of mortar should be changed. The robot may self-organize, and allocate the required steps with other robots. Alternatively, or in combination, the robots may communicate with a central computer which orchestrates the robots.

In a preferred example, the robots is configured to: * Pick up a brick * Pick up a wall tie * Fill with mortar * Take a step along (traverse) * Take a step up (climb up) * Take a step down (climb down) * Make a turn through any angle * Put a wall tie into position * Apply mortar to the existing structure * Place a brick * Fine tune the position of the brick after placement * Point the mortar and brush the mortar Robot Body, Legs, and Feet The robot comprises a robot body. The robot body is where the majority of the motors, control boards and batteries may be situated. Motors may use belt drives or other mechanical linkage to transmit movement to the robot's extremities.

The robot comprises two legs which are used for manoeuvring the robot. The legs are each configured to move in a movement plane, horizontally and vertically relative to the robot body, in a linear fashion. The legs slide along bearings which interface with the main chassis. The legs are mounted on opposite sides of the chassis to allow the legs to pass each other when taking a step. The legs are spaced wide enough apart to allow a structural unit to pass between them as the legs pass each other. Alternatively, they may be mounted closer together, but include a wider opening between them at some point over their length to allow a structural unity to pass between them.

The robot comprises a foot at the end of each leg. The feet are mounted so that the underside or 'sole' is horizontal. At least part of each foot extends from the leg to which it is coupled towards the opposite side of the robot body. In other words, the sole of each foot extends in a horizontal direction from the leg to which it is coupled towards the movement plane of the opposite leg. The feet are able to rotate around a vertical axis. In some embodiments, the feet are configured to pivot in the other two axes, to mitigate for uneven ground or non-horizontal structure surface. The feet can have varying undersides or 'soles' depending on what structural unit the robot is assembling a structure with. For example: flat for flat bricks, a round protrusion to insert into bricks with round holes, a square protrusion to insert into bricks with square holes, or a sculpted protrusion to fit into a frogged brick. In some embodiments, the feet or feet soles are interchangeable with simple fixings or snap fit etc. Figure 1 shows a close-up diagram of an exemplary robot 1000.

It consists of two legs 1010 1011 that attach to the main chassis 1020 of the robot 1000 and move horizontally and vertically relative to it in a linear fashion. That is, they slide along bearings which interface with the main chassis 1020. They are mounted on opposite sides of the chassis to allow them to pass each other when taking a step. They are spaced wide enough apart to just allow a brick to pass through.

At the bottom of the legs are the feet 1030 1031. The legs attach to the feet through a joint that is able to rotate about the vertical axis. This is illustrated by the cylindrical bearing between the leg and the base of the foot. This enables the robot to rotate the main chassis when standing on one leg. The feet are mounted so that the 'sole' is horizontal. Future additions could allow the feet to additionally pivot in the other two axes, so they could mitigate for uneven ground or non-horizontal structure surface. The feet can have varying undersides or 'soles' depending on what brick the robot is constructing with, for example: flat for flat bricks, a round protrusion to insert into bricks with round holes, a square protrusion to insert into bricks with square holes, a sculpted protrusion to fit into a frogged brick. These could be easily interchangeable with simple fixings or snap fit etc. Between the two legs and attached on the underside of the main chassis is the end-effector or gripper 1040. This moves horizontally along the main chassis independently of the two legs. It has a clasping mechanism (a set of jaws 1050) which can open or close to grip and hold a brick. The left-hand jaw 1050 is seen as the rectangle protruding down from the bulk of the gripper. The jaws grip the long outer edges of the brick to allow smaller bricks to also be held when required. If perforated or frogged bricks are being used, the gripper may also hold the brick from the inside.

The two cylindrical puck shaped objects 1060 1061, one on the top of the main chassis and the other rotated 90 degrees at the head of the main chassis are the LIDAR scanners. These scan 360 degrees in a circular motion generating point-cloud data of the brick structure and surrounding environment. This data may be used as a record of the as-built structure as well as for localisation using Simultaneous Localisation and Mapping (SLAM).

At the front, hanging down from the underside of the main chassis is the mortar nozzle 1070. Above this on the top side of the main chassis is the opening 1080 through which the mortar dispenser is able to fill the mortar storage canister that is inside the main chassis 1020. The two semi-circular objects that form the opening slide along the rectangular objects either side to reveal an opening. The nozzle 1070 on the underside is able to extend, rotate about a vertical axis, and apply the mortar to the area of the wall where the brick is to be placed.

Main Effector The robot further comprises a main effector, such as an end effector. The main effector may be a main gripper. The main gripper is configured to grip structural units, such as a brick.

The main gripper moves horizontally along the main chassis independently of the two legs. It is mounted on the underside of the chassis, between the paths of the two legs. It has a clasping mechanism (for example, a set of jaws) which can open or close to grip and hold a structural unit. The clasping mechanism grips the long outer edges of the structural unit. By using a clasping mechanism which grips the outer edges of a structural unit, the same clasping mechanism may be used to hold structural units of different widths by opening and closing the clasping mechanism according to the width of the brick to be held. If perforated or frogged structural units are being used, the main gripper may also hold the structural unit from the inside.

Figure 8 shows a perspective view of the exposed workings of an exemplary robot gripper 82040, shown clasping a typical brick 82090. The mechanism is composed of a main plate 8250, which is approximately square. This is joined to the rest of the robot's mechanism via a bearing joint 8260 which allows the whole gripper assembly to rotate around the vertical axis. A central hole within this bearing allows wires to pass to the rest of the gripper mechanism while it rotates. A linear actuator is mounted in the middle of the plate 8270, controlling the movement of the two gripper jaws 8050-1. The jaws can move horizontally in a linear motion, moving closer together or further apart. The jaw ends consist of a vertically mounted flat plate 8050-1 which interfaces with the side of the brick. The plate may have a shaped or stepped end profile to allow it to also interface with the top of the brick, thus calibrating the brick's vertical and lateral position. At each corner of the plate is a sensor see figure 9 to measure distance between the top of the brick and the plate. Arranged in symmetrical pairs, there are tapping or vibration mechanisms 8280 to adjust the relative position of the brick once it has been placed on to the mortar bed and released from the gripper jaws. Each mechanism consists of a motor 8290-9 with an eccentric mass attached to its shaft. A Linear Resonant Actuator type vibration motor may also be used. The motor assembly is mounted on a linear element, connected at one end with a hinge joint to the main plate. At the end of the linear element is a shaped, dense, and hard-wearing element to allow good transfer of force to the brick, like a hammer head 2300-9. When the motor is made to turn, the eccentric mass induces the linear element to move through its arc of motion when tethered by the hinge joint. At one end of the arc, it may be supported by a spring restriction to increase the return force. At the other end of the arc, it makes contact with the surface of the brick, transferring a high impact force. When these are used, it is assumed no or minimal mechanical contact between the gripper assembly and the brick, the only contact being when the tapping mechanisms deliver their impacts. The tapping mechanisms may be brought into tapping range of the brick by widening the gripper jaws to allow the main plate to be lowered. The tapping mechanisms are mounted such that they can deliver impacts to the brick edges from each side and from the top. In the example shown, there are two on each brick side edge, two on the end, and four to affect the forces to the top of the brick. The gripper can be moved around by the robot's other degrees of freedom such that the exact impact locations of these mechanisms can be finely tuned. A control program interprets the distance sensors in combination with other sensors on the main robot body such as cameras, LIDAR, accelerometers etc. to determine which tapping mechanism should be applied, where precisely on the brick it should be applied, and how large the impact force should be.

Figure 9 shows a perspective view of the underside of the exemplary robot gripper 9040, shown without a brick in place. This shows the aforementioned elements: the flat underside of the main plate 9250, with a distance sensor 9310-9 in each corner pointing downwards towards the brick face. The gripper jaws 9050-1 are free to move horizontally, and feature a shaped or stepped end profile to interface with the sides and top of the brick. The linear elements of the tapping mechanisms with their respective end pieces are allowed to access the brick from the sides or top. The mechanisms with affect the top of the brick 9301 9303 9307 9038 protrude through hollows in the main plate.

Connecting material Storage and Delivery System The robot may include a connecting material storage and delivery system. The connecting material storage and delivery system is sized appropriately to store at least enough connecting material to lay a single brick. The connecting material storage and delivery system comprises a connecting material container attached to the robot body. The connecting material container is connected to a section comprising a shaped opening. The section comprising a shaped opening is referred to herein as a nozzle. The connecting material container may include a second opening for filling the connecting material container, however in some embodiments the connecting material container is filled via the nozzle.

In an illustrative example, the connecting material storage and delivery system comprises a cylindrical connecting material container mounted vertically within the robot body. The connecting material container comprises an opening in the top and a shaped opening in the bottom (the nozzle).

The nozzle may be detachable element (e.g. a snap or screw fitting), that is detachable from the remainder of the connecting material storage and delivery system, to allow nozzles with various shapes to be fitted to the connecting material storage and delivery system. This provides for shaping of the connecting material into different application shapes as the connecting material exits the nozzle as required for different brick types. This may also allow easy assembly, maintenance and cleaning of the nozzle and connecting material storage and delivery system. The nozzle opening may be angled to allow connecting material exiting the connecting material storage and delivery system to adhere to both horizontal and vertical surfaces. In a preferred embodiment, the nozzle opening is angled at 45 degrees.

An example application shape for the connecting material is a rectangle slightly taller and slightly narrower than the final volume of connecting material once it is in place with a brick on top of it. This application shape allows the connecting material to be compacted into final position, generating good surface adhesion, without creating large excess needing to be cleaned up afterwards during pointing. When the connecting material is mortar, this system greatly reduces the amount of mortar used when compared with traditional methods of brick laying wherein the mortar is placed on the structure by hand. Alternative shapes may also be used, such as a rectangle with rounded corners, a cylinder, or a rectangle with V-shaped projections.

The nozzle may have a nozzle cover which covers the opening of the nozzle. For example, a flap, or a set of flaps, which cover the opening of the nozzle. This nozzle cover allows the connecting material storage and delivery system to build up pressure behind the nozzle cover, such that the connecting material may then be released in a smooth stream. The nozzle cover could be passive e.g. spring loaded, or active e.g. controlled by an actuator.

The top of the connecting material container may be covered by a moveable cap. The moveable cap may be opened by the application of force to the moveable cap. The force may be applied by an actuator within the robot. Alternatively, the force may be applied using something outside the robot -for example, the force may be applied by the robot pushing itself into something, or by something pushing itself into the robot. The moveable cap may be closed by the application of force. Again, the force may be applied by an actuator within the robot, or something outside the robot. Alternatively, the moveable cap may be spring loaded to snap shut after the force which opened the moveable cap is released (e.g. the cap is a like a sliding door with sloping tops, or a trapdoor which opens downwards). In a preferred embodiment, the moveable cap is configured to open when a vertical force is applied to the cap. the robot is configured to 'crouch down' with the moveable cap positioned underneath the nozzle of a connecting material supply point and then 'stand up', thus interlocking the nozzle of the connecting material supply point with the opening of the robot's connecting material container. The connecting material supply point may then supply the robot with connecting material.

Inside the connecting material container, there may be a mechanism for moving the connecting material towards the nozzle. In a preferred embodiment, this mechanism is an auger screw which is sized to fill a cylindrical connecting material container. As it turns (powered by a motor), it pushes connecting material down the container and expels it out the end, under pressure and at a controllable rate. One alternative is to use a piston which moves down the length of the tube and similarly expels the connecting material under pressure and at a controllable rate.

The connecting material storage and delivery system may comprise a Connecting materialflow control system which controls the flow of connecting material from the nozzle. The flow control system improves the accuracy of the flow of connecting material from the nozzle by measuring the flow of connecting material from the nozzle and adjusting the mechanism for pushing connecting material through the nozzle. The flow control system comprises sensors are used near the nozzle opening to determine the rate of connecting material flow. The sensor may comprise a flow-meter, such as an optical flow-meter or a paddle wheel meter to measure the flow. Alternatively, sensing can be avoided by calibrating the system which pushes the connecting material through the nozzle. For example, an archimedes screw may control the rate of flow. Some connecting material may be deliberately expelled in advance of the desired application, such that the nozzle is fully filled with connecting material and any air pockets have been removed.

The nozzle may be configured to rotate around a vertical axis. This allows the end of the nozzle to be controllably positioned such that when the connecting material exits the nozzle, it can be applied to the vertical faces of the structural units already in place to the side(s) of the intended location of the structural unit to be placed.

The nozzle may be configured to move horizontally relative to the robot body in order to apply connecting material along the horizontal top surface of the structural unit below the intended location of the structural unit to be placed.effector. However, the number of moving parts may be reduced by fixing the horizontal position of the nozzle relative to the robot body, and configuring the robot to move its body, and thereby the nozzle, horizontally in order to apply connecting material along the horizontal top surface of the structural unit below the intended location of the structural unit to be placed.

The nozzle may not extend vertically from the robot body as far as the legs when the robot is body is in its tallest position such that the nozzle is clear of the structure when the robot manoeuvres around the structure. In this case, the robot is configured to lower its body, and thereby the nozzle, towards the structure when applying connecting material. Alternatively, the nozzle may be configured to move vertically relative to the robot body. Of course, a robot may use both methods.

The connecting material storage and delivery system may have a depth sensor or other sensor to assess how much connecting material the system currently contains. This can inform the robot as to the amount of mortar required to refill the connecting material storage and delivery system for the next given brick to lay, and the length of time required for refilling. The robot control program may be configured to adjust the amount of connecting material required depending on the brick type and position in the structure.

The connecting material storage and delivery system may include at least one sensor for checking the moisture content of the connecting material. In such a system, the robot may alert the controller if the mix is not correct. Alternatively, the robot may add water from an onboard reservoir as required. The robot may also comprise a reservoir for storing an additional chemical, such as a catalyst agent, which is added to the connecting material just as the connecting material is expelled from the nozzle, to make the connecting material set more quickly.

In the case where additional water or chemicals are added to the connecting material, the robot may use a mechanism to mix the connecting material within the connecting material storage and delivery system. This may be a single or multiple rotating paddle wheels located within the container. Alternatively, the connecting material may be passed through a series of fins or constricted openings to passively mix the connecting material.

The nozzle may have a specially designed pointing devicefor pointing the connecting material after the structural unit has been laid. The pointing device may include a pointing element integrated into the form of the nozzle or be a separate detachable and interchangeable pointing element. The use of separate detachable and interchangeable pointing elements allows for different pointing styles such as square inset, raked, rounded etc. In order to point the connecting material, the robot may move the nozzle through a motion similar to that performed when applying connecting material, but in this case extend the nozzle and scrape the pointing element along the external facing surface of the connecting material joint. The excess connecting material may be scooped up into a temporary storage point built into the nozzle. Alternatively, the excess connecting material may be deliberately scraped off and encouraged to fall to the ground.

The nozzle may have brush element attached to it e.g. on the reverse side of the nozzle.

This allows the robot to brush the face of the newly laid brick and do a final clean-up of the connecting material after it has been pointed. In order to brush the connecting material, the robot may be configured to move the nozzle in a similar way to the way the robot moves the nozzle when it applies the connecting material, but this time it reaches down the outside of the new brick which has been laid and passing the brush element along the connecting material bead.

In another embodiment, the connecting material may be prepacked into cartridges of appropriate size. These cartridges may hold the connecting material within a single use or reusable container, for example made from a thin plastic film. In this embodiment, the cartridges are loaded into the storage and delivery system as sealed units. They may then be squeezed under pressure to puncture the cartridge and expel the connecting material from the nozzle. Alternatively, the cartridge may employ a valve to interface with the connecting material storage and delivery system. This would allow the connecting material to flow from the cartridge only after the valve has been opened by the storage and delivery system.

Figure 4 shows a perspective view of an example connecting material storage and delivery system 4100. This consists of a hollow cylinder 4110 in which the connecting material is stored, here shown with a cutaway view. A cap 4120-1 at the top of the cylinder allows connecting material to be poured in through an opening 1080. The cap 4120-1 is shaped with an indentation such that when a force is applied to it from the top, for example by a round tube seeking to enter the hole, then the sides of the cap are pushed apart to reveal the opening. The moveable pieces of the cap 4120-1 move along rails 4130-1 and are pushed back by springs, such that when the tube is removed from the opening, the cap automatically springs back shut. Below this cap, there is a mounting point 4140 by which the cylinder is attached to the main robot chassis. This mounting point may be moved vertically or rotated, such that the whole cylinder can be lowered from the robot and rotated to direct the flow of connecting material from the nozzle. Built into this mounting, there is a motor 4150 which is connected to a rotating shaft 4160 contained inside the hollow cylinder, using a mechanical linkage such as a belt. The shaft is connected to an auger screw 4170, sized to abut the edges of the cylinder. Thus when the screw is rotated, and connecting material in the tube is moved down the cylinder's length. A flow sensor in the mouth of the nozzle see Fig 5 5200, combined with changing the auger's speed of rotation allows the connecting material flow to be controlled. At the end of the cylinder, there is a nozzle 4070 to direct the connecting material out of the cylinder in a desired formation. This nozzle assembly may be a snap or screw fitting 4220, such that it can be removed for cleaning or changed to give an alternative connecting material flow shape. On the sides or edges of the nozzle fitting there are attachments for pointing the connecting material, and a brush for cleaning the connecting material. See Figure 5.

Figure 5 shows a perspective view of an example nozzle 5070 fitting attached to the end of the connecting material container cylinder 5110. It shows the main body of the cylinder, with a depth gauge sensor 5180 to signal the level of connecting material left in the container. The nozzle assembly attaches to the cylinder using a screw or snap fitting 5220-1, to allow cleaning, maintenance or alternation of nozzle shape. The nozzle opening is rectangular, at 45 degrees to the vertical plane, to allow adhesion of the dispensed connecting material to vertical and horizontal surfaces. There are curved or angled fins 5190 fitted inside the nozzle, shaped in such a way to encourage the connecting material to fill the whole of the nozzle exit and assist its transition from the round cylinder 5110, propelled by the auger screw 4170, into the square opening. A flow sensor 5200 mounted near the nozzle opening allows monitoring of the current rate of connecting material flow. At each side of the nozzle, there are additional fittings to allow pointing of the connecting material. These are mounted at an angle to allow use in the vertical and horizontal directions without adjustment. This example shows a curved projection, proportioned so that when it is passed across the connecting material joint of a freshly-laid brick, it leaves a desirable concave pointing style. Alternatively, a different fitting could be designed to give straight or angled pointing style.

These fittings are detachable e.g. by snap fit or threaded connection, so they can be easily cleaned or changed. On the rear of the nozzle, there is a brush fitting. This is designed to allow the pointed connecting material joint to be brushed free of small flecks of connecting material as is customary. The brush attachment may also be snap fit to allow easy maintenance.

Figure 13 shows twelve diagrams illustrating the intermediate configurations of the exemplary robot during the brick placement action. The illustrations are ordered from left to right, starting at the top and working down. Frame 1 shows the starting configuration of the robot standing on the brick structure in front of the area where the brick is to be placed.

Frame 2 shows the extended nozzle ready to spread the connecting material. Frame 3 shows the starting position for the connecting material spreading where the main chassis of the robot has been lowered and moved backwards to put the connecting material nozzle in the lower right hand corner of the brick placement region. Frame 4 shows the end of the first connecting material spreading action. The main chassis of the body has been raised, lifting the connecting material nozzle as the connecting material is pumped out, spreading a layer of connecting material on the head of the brick next to the placement region. Frame 5 shows the starting position for the next phase of connecting material spreading. The main chassis of the body has moved forward in front of the connecting material already spread in the previous phase as well as lowered to its starting position. Frame 6 shows the final robot configuration after the connecting material spreading. The main chassis has been moved forward as the connecting material nozzle extrudes connecting material over the area where the brick is to be laid. Frame 7 shows the robot in its starting configuration as was shown in Frame 1. The nozzle has been retracted and the main chassis raised and pulled back. Frame 8 shows the robot configuration just before laying the brick. The gripper has moved forwards to place the brick above the region where it is to be placed. Frame 9 shows the configuration of the robot as the brick is placed. The main chassis has been lowered with respect to the legs putting the brick on top of the connecting material. Frame 10 shows the robot returning to its configuration having placed the brick. The gripper has released the brick in its position, making all necessary finer adjustments to the bricks position using the tapping and sensing mechanism, and has raised the main chassis back to its walking height. Frame 11 shows a diagram of the configuration of the robot during pointing. The rear leg has raised, allowing the main chassis of the robot to be rotated so that the connecting material nozzle with its pointing attachment can reach the side of the brick. The main chassis is then able to move back and forth, whilst rotating, brushing the side of the brick and removing any excess connecting material. Frame 12 shows the robot returned to its initial configuration, with the brick placed.

Figure 10 is a diagram of the exemplary robot 10000 configuration just after it has spread the connecting material for a brick 10090 to be laid. The robot is located on top a brick structure. The nozzle 10110 is extended and the main chassis 10020 is lowered so that the head of the nozzle is positioned just above the area where the connecting material is to be spread.

The brick 10090 is clasped in the gripper waiting to be placed. The connecting material can be seen already spread on top of the brick structure in the region where the brick is to be placed.

Figure 11 is a diagram of the exemplary robot 11000 configuration just before placing a brick 11090 on a brick wall. The connecting material is illustrated spread over the region where the brick is to be placed. The gripper is positioned over this region with a brick clasped between its jaws.

Figure 12 is a diagram of the robot configuration as it places the brick in the connecting material. The main chassis 12020 has been lowered such that the brick is placed on top of the connecting material bed. It is at this point that the tapping mechanism in the gripper makes final adjustments to the bricks position, once the jaws of the gripper have been released. This detail is not shown in the diagram.

Wall tie applicator The robot may comprise a wall tie applicator. The wall tie applicator comprises a small wall tie gripper mounted on the rear of the connecting material nozzle or on a leg or the main gripper. The wall tie gripper can clasp and rotate wall ties. Wall ties are configured to be attached to an inner structure. For example, by drilling or nailing them into the inner structure. In a preferred embodiment, the wall ties are configured to fit into extruded metal U-sections attached to the inner structure. The wall ties may be inserted into the U-section while in the vertical plane. The wall ties may interface with the U-section such that once the wall tie is inserted, it cannot be removed. For example, the wall tie may comprise teeth that engage with a pawl in the U-section to form a ratchet, so that as the end of the wall tie is inserted in the U-section, it cannot come undone. Alternatively, the wall tie may be rotated to latch into the U-section. The wall ties may have small notches to interface with the U-section.

In another scenario, the assembly system comprises a dedicated 'wall tie' robot, which specialises in attaching wall ties. The dedicated wall tie robot comprises a specialised wall tie effector for the main effector, in place of the main gripper, which is configured to store and manipulate wall ties, and be able to place, drill, and/or nail them into an existing inner structure.

Charging The robot may comprise a charging point for charging the robots battery. The charging point may be configured to allow for an automated charging at a charging station. The charging may be contactless through induction. The charging point may be on the robot body. Alternatively, the charging point may be integrated into the robot's feet to allow charging by standing the robot on a certain point (or charged continuously as it walks over certain parts of the structure such as the staircase explained below). The robot may also have other sources of power such as solar panels.

In the case where the charging point consists of exposed electrical contacts, a cover may be used to protect them from moisture or dirt. When charging is desired, this cover may be actively moved by an actuator, or passively pushed aside by the robot, so as to expose the contacts to the charging station.

Figure 32shows an example of a unit designed to allow the robots 32000 to charge their batteries autonomously 32960. It consists of a number of docking stations mounted on a single unit 32970, although these docking stations could also be separated into distinct units themselves. It may have a carry handle or wheel to allow easy transportation. This single unit is a portable structure containing a computer controller and means to dispense electrical power to the robots' batteries. It may have larger internal batteries, or be connected to an external power source 32980 such as a solar panel, wind turbine, or generator etc., or mains power cable 32990. The solar power and mains power cables are illustrated in the diagram. The unit has male elements which engage with the robot's batteries, with contact or induction charging capabilities. This male element is designed to engage with the battery and hold it in position when not connected to the robot. Pathways 32240 for the robot, consisting of a designed surface suited to interface with the robot's foot allow the robot to approach the connection points on the unit. These pathways are connected or placed next to robot's main work area to allow easy access. The pathways may have virtual checkpoints to communicate with other robots that a certain charging bay or pathway is occupied. By controlling its movements, the robot can hook its battery onto the male element on the unit and allow it to charge. Alternatively, after the battery is attached to the base unit, the robot may uncouple from the battery and leave it charging. The robot can then proceed using the designed pathways to another available battery on the unit, which has been previously charged. The robot may be powered during this intermediate time, while it has no main battery connected, using another internal battery, or contact or induction power via its feet on the pathways. Comms

The robot may also integrate aerials or receivers for wireless communication with other robots or location systems via VVi-Fi, Bluetooth, laser or infra-red etc. Balancing The robot's various degrees of freedom (position and rotations of legs, gripper, nozzle and chassis) are controlled and coordinated by the robot control program such that the overall system is always balanced. The robot knows what type of structural unit and how much connecting material it is carrying at any given time, so adjusts its balancing to suit. For example, when taking a step, the robot control program arranges the robot body and main effector so the centre of mass is over one leg; then it lifts the other leg; then it moves the raised leg past the stationary leg whilst the robot body and/or main effector are moved in the opposite direction at the required speed and acceleration to maintain a net zero jerk force on the robot; then it places the leg down and repeats to take another step.

In one example, each motor has at least one rotary encoder attached to its output shaft. This allows the position, speed and acceleration of the shaft to be determined at any point in time.

This allows a Proportional Integral Derivate (PID) control algorithm to be implemented. The motor's encoder may be connected to a central controller, or to a decentralised controller for each motor. In the decentralised case, a central controller may give only top-down commands to each motor, and each motor controller implements the PID algorithm to achieve the desired result.

Net zero jerk force is maintained by treating each primary component of the robot assembly as a movable mass. For example, the main beam may weigh 5kg, and each leg may weigh 1kg, and the gripper may weigh 3kg. If the gripper stays in the same place relative to the main beam (e.g. during a typical step), the total mass of the main beam and the gripper would be 8kg. Using these component weights we can take the example of a robot balanced over one of its legs, leg A. The other leg, leg B, may, for example, have its centre of mass 8x distance from leg A on one side. In this case, for the robot to be balanced, the main beam and gripper must have their centre of mass x distance away from leg A on the other side.

During the stepping motion, leg B moves past leg A to be 8x distance away from leg A on the opposite side from where it started. Similarly, while this movement occurs, the main beam and gripper are moved such that they end up x distance away from leg A on the opposite side from where they started. These two motions are controlled to take the same amount of time. The result is that Leg B and the main beam and gripper can be seen to swap sides relative to the stationary leg A, and the main beam and gripper are moved with 1/8th the speed and acceleration of leg B. Figure 2 shows twelve exemplary diagrams illustrating the intermediate configurations of the robot 2000 joints during a step-up action onto a brick. The illustrations are ordered from left to right, starting at the top and working down. Each diagram shows a line of five bricks on top of which the robot is positioned. Frame 1 shows the robot in a standing position. Frame 2 shows the robot moving the gripper and chassis, illustrated holding a brick, move forwards so that the centre of gravity of the robot is over the front leg. Frame 3 shows the robot lifting its rear leg and balancing on the front leg. Once the leg is lifted the robot moves the leg forwards, maintaining balance by moving the gripper and chassis at the correct speed in the opposite direction. Frame 4 illustrates the mid-point of this manoeuvre, where the gripper, chassis, raised leg and fixed leg are aligned, balancing the robot. When carrying connecting material this mid-point arrangement may be different as the mass distribution of the robot changes. Frame 5 shows the configuration of the robot when the raised leg has reached the intended position ready to be placed down. Note the gripper and chassis have moved further back to compensate for the weight of the leg at the front of the robot. Frame 6 in a standing configuration with the raised leg now placed down in its intended position, completing the first step. Frame 7 shows the configuration of the robot after the gripper and chassis of the robot have moved forward to place the centre of mass over the front leg once again, achieving the same configuration as Frame 2 but having switched the places of the right and left leg. All subsequent frames show the same configurations as for the corresponding frame in the previous stepping action but with the right and left leg switched, (e.g. Frame 8 <-> Frame 3, Frame 9 <-> Frame 4). Frame 8 shows the rear leg raised, balancing the robot on the front leg. Frame 9 shows a still mid-step as the raised leg moves forwards and the gripper moves back to keep the robot balanced. Frame 10 shows the end of the stepping action with the raised leg above its intended placement location. Frame 11 shows the robot in a standing configuration with the raised leg now lowered. Frame 12 shows the final standing configuration, identical to the configuration of Frame 1 but with the robot now two brick lengths further along the brick structure it is standing on, having completed two steps.

Figure 3 shows twelve diagrams illustrating exemplary intermediate configurations of the robot joints during a step-up action onto a brick. The illustrations are ordered from left to right, starting at the top and working down. Each diagram shows a line of five bricks on top of which the robot and one other brick are positioned. The frames correspond exactly to the frames described for the Figure 2 figure but with a differing final height for the leg taking the first step as it climbs on the brick.

The robot may comprise at least one robot balancing system. The robot balance system enables the robot to balance on one foot in order to take steps and rotate. For the robot to balance correctly it needs to either know the mass distribution in the robot, which regularly changes, or use sensors to detect when it is off balance and adjust dynamically. Several methods may be used in conjunction. For example, the planned mass distribution method (previously described) may be used as a coarse grain method, with a sensor method used to fine-tune or safeguard the robot's balance in finer detail. Three exemplary robot balance systems for ensuring the robot is balanced using sensors are described below.

The robot may comprise a pressure robot balancing system. A pressure robot balancing system comprises pressure sensors in the four corners of the feet of the robot. If the robot is correctly balanced and the surface the robot is standing on is flat, then the pressure at all four corners of the foot of the robot should be equal. The pressure robot balancing system comprises a control loop which monitors the pressure at a plurality of points on the robot-for example, the at the feet of the robot -and controls the position of the main effector to continually balance the robot. By monitoring the pressure at, for example, the robot's feet, the pressure robot balancing system can quickly adjust the position of the main effector to minimise the difference in values reported by the pressure sensors. For example, for a robot with pressure sensors in its feet and arranged in a pose such that it is standing on one leg, if the pressure robot balance system detects a pressure gradient across this foot along the axis of the main beam, the robot corrects by moving the gripper or effector along the main beam in the direction of decreasing pressure. The speed and exact manner of the response by the gripper are controlled by, for example, a PID controller.

The robot may comprise a strain robot balancing system. A strain robot balancing system comprises strain gauges -for example in the legs and main beam of the robot -which can be used to detect bending and flexing that will occur as the robot tips. The strain robot balancing system comprises a control loop configured to monitor the strain gauges and control the position of the main effector to continually balance the robot. The control loop may be configured to minimise strain. For example, for a robot with strain gauges in its legs and arranged in a pose such that it is standing on one leg, if the strain robot balance mechanism detects a strain in its standing leg in a particular direction it corrects by moving the effector or gripper in the opposite direction, thereby reducing the strain. The speed and exact manner of the response by the gripper are controlled by, for example, a PID controller.

The robot may comprise a gyroscopic robot balancing system. A gyroscopic robot balancing system comprises at least one gyroscopic sensor to detect changes in angular momentum. Certain changes in angular momentum indicate that the robot is losing balance. When the gyroscopic robot balancing system detects such changes in angular momentum, the robot adjusts the position of the main effector to balance the robot. The gyroscopic robot balancing system comprises a control loop configured to monitor changes in angular momentum and control the position of the main effector to continually balance the robot. For example if the gyroscopic robot balance system detects a tilt in a particular direction along the axis of the main beam it corrects by moving the effector or gripper in the opposite direction, thereby reducing the detected tilt. The speed and exact manner of the response by the gripper are controlled by, for example, a PID controller.

The control loops of the robot balancing systems may also be configured to control the position of other parts of the robot to maintain balance. For example, the position of the main beam or the spatial arrangement of connecting material within the connecting material container.

The control loops of any robot control system may use proportional, integral, and/or differential control.

Collision avoidance In order to ensure safety and avoid accidents the robot may include at least one collision avoidance system. The collision avoidance system detects potential collisions. Upon detection of a potential collision, the robot will pause before calculating evasive or alternative actions. Two exemplary collision avoidance systems are described below.

The robot may include a proximity collision avoidance system. In a proximity collision avoidance system, proximity sensors placed at key points along the main beam of the robot. Since all joints of the robot move relative to the robot body, all collisions can be avoided by ensuring that the areas around the robot body and the space below it are free from obstructions. A LI DAR sensor scanning in a plane perpendicular to the robot body is able to detect potential obstructions as the robot manoeuvres around, covering the area beneath the main beam. The proximity sensors ensure the robot body avoids collisions by pausing the robot actions if the values fall below a safety threshold.

The robot may include a stereoscopic collision avoidance system. A stereoscopic collision avoidance system may comprise two stereoscopic wide-angle cameras, one facing a first direction of robot travel, and the other facing the opposite direction of robot travel. Using stereoscopic vision enables a depth-map to be obtained for the areas within the camera's field of view. Using the depth-map, distances to objects are obtained and threshold values can be set below which the robot can pause operations to avoid collisions. Stereoscopic cameras have the added benefit of being able to detect and identify particular environmental objects, e.g. humans, and adopt larger safety thresholds to take into account the behaviour of the detected object, e.g. unpredictable human behaviour, ensuring greater safety.

Assembly control program The automated assembly system may comprise assembly control program. The assembly control program orchestrates the robots. The assembly control program may control each robot by configuring each robot's individual robot control program before each structure is assembled. Alternatively, the assembly control program may control the robots by continually communicating with each robot's individual robot control program during the assembly process.

The key input to the robot organisation program is a model of the design of the structure that is to be built. The construction industry has recently adopted Building Information Modelling (BIM) as the standard method for designing and digitally representing buildings. To seamlessly fit in with current practices in the construction industry the robot organisation program may be configured to use BIM models.

BIM models of buildings typically represent brick structures by their dimensions, e.g. height, width and depth, including the dimensions of cut-out regions for windows etc. By specifying a brick-type and bond, a structural unit layout model is generated from the BIM specifications. The structural unit layout model provides a plan including the 3D location and orientation of each structural unit in the structure. For example, the structural unit layout model may provide the position and orientation of each complete brick and each cut-brick, the total number of bricks required, and the total number of courses.

Once the structural unit layout model has been determined from the original BIM model, the robot organisation program may calculate the build-time and costs for assembling the structure as a function of the number of robots and different material supply locations. Optimal solutions can be computed by defining the available space on-site around the structure, which may constrain the possible material supply locations.

The state of the structure at all stages of the assembly process is determined by the assembly control program, such that the structure is always traversable with the robot's capabilities. For example, if a robot is only capable of stepping up or down the height of one structural unit, then the assembly control program plans the assembly process such that there is no need to make a robot step up or down more than the height of one structural unit.

The assembly control program may determine the number of robots and additional hardware required for assembly. Alternatively, the assembly control program may determine the assembly process constrained by the number of robots and/or the use of certain additional hardware.

Due to the flexibility of the robot organisation program, the structural unit layout model can be recomputed late-on into the build process. If the as-built structure deviates from the BIM design (e.g. due to mistakes being made or late design changes), these deviations can be taken into account and all the above mentioned properties can be recomputed. This process can be automated using 3D sensing technology, such as LIDAR, and machine learning algorithms to match the sensed 3D data, such as point-cloud data, to the corresponding BIM components. By scanning the as-built structure, real world measurements can be compared to the BIM design. By updating the BIM model with as-built dimensions, any necessary change to the brick arrangement due to deviations from the design can be computed. In this way the robot organisation program is able to control the robots to make adjustments to the structure in a way that human bricklayers routinely do.

All the data collected during the build process (e.g. 3D scans, photograph or video) can be stored as record of the construction process. This is a useful resource if future alterations need to be made, or for insurance/liability purposes. All of the data can be loaded back into the BIM data environment (including time, weather conditions, material origins etc of each brick laid).

Figure 18 illustrates at a high level the software process taking a BIM model design of a structure all the way through to the actions of the robot. The BIM analysis flow diagram on the left illustrates the process of extracting the information required by the robots from the BIM model. Starting with a BIM model as input 18410 as well as extra information about the brick type and the bond type, i.e. how the bricks are arranged, the software calculates the position and orientation of each brick 18420. This can be done automatically by comparing the dimensions of the chosen brick type to the region of space they must fill as dictated by the BIM model. By adjusting the connecting material gap and the brick sizes e.g. cut or full, the software can find the optimal arrangement to fill the required space. The next step calculates the optimal number of robots, supply material arrangement and equipment 18430 based upon the structure and the number of bricks as calculated in the previous step. This can be augmented with extra information, such as available space around the build site and desired costs/time constraints. Final once the layout of the supply materials has been fixed the software calculates the necessary flow control for the robots, such as valves and checkpoints. The valves and checkpoints may ensure an efficient and safe building process 18440.

The diagram then illustrates that the information calculated from this analysis is passed on to the robot 18450. In particular, the brick positions so that the robot knows where to place what bricks, the order for laying the bricks so they know which brick to lay when, the locations of the supply materials so that they can navigate and pick-up supply/charge their batteries etc, and the flow control arrangements so that they can obey the rules correctly. The high level control loop for the robot is illustrated in the "robot analysis". The initial step is deciding upon the next action depending on the state of the robot and the state of the build18460 18470. Once this action has been decided the joint movements for the action are calculated 18480. These could be the joint movements for stepping to the location on the build site that the robot is required to reach or the actions for placing a brick etc. For the former case the robot takes into account any movement restrictions imposed by valves that control the robot flow 18490. Once these are calculated the joint movements are iterated through one by one, each time checking for collisions and checkpoints18500 18510. If it is safe to proceed the joint movement is carried out and this repeats until the action is complete.

Robot Positioning Each robot may comprise at least one positioning system which determines the global position of a feature of the robot. The robots may have multiple positioning features so that if one fails there is always a back-up system. From the global position of the positioning feature, the position of each component of the robot can be determined by determining the relative position of the robot components with respect to the positioning feature. The robots movements may be calibrated such that even after the robot has moved, using knowledge of the movements made the robot can determine the relative position of the robot components with respect to the positioning feature.

The robot may include a motive positioning system for determining the robot's position on the structure. A motive positioning system calculates a change in the robot's position based upon the movement of the robot. The distance and direction of each robot step will be accurate to within the margin of error associated with the precision of the joint movements.

From a known starting position, the robot's position on the structure can be computed from the series of actions of the robot.

The robot may include a count positioning system for determining the robot's position on the structure. A count positioning system comprises a sensor for detecting the structural units around the robots feet. For example, a LIDAR sensor configured to scan the bricks around the robots feet. The LI DAR sensor may be positioned such that it scans in a circle perpendicular to the directions in which the robot manoeuvres when traversing the automated assembly system, for example. The LIDAR sensor may be positioned, for example, on the front of the robot. As the robot manoeuvres, it builds up a 3D model, such as a point cloud dataset, of the structural units, such as bricks, structure, and/or other components of the automated assembly system beneath the robot. Using this model, the position of the robot relative to the structure can be determined by calculating, for example, the number of structural units traversed. Detecting distinct structural units in a model, such as a point cloud dataset, can be achieved using a convolution neural network or other machine learning algorithm. The convolution neural network or other machine learning algorithm is trained for detecting distinct structural units in a model.

The robot may include a simultaneous mapping and localisation system. The simultaneous mapping and localisation system is configured to determine a map of the structure in the locality of the robot and the robot's location in that map. The simultaneous mapping and localisation system comprises at least one sensor for mapping the robots environment. The sensor may be a LIDAR scanner. An in-plane LIDAR scanner can be placed, for example, on top of the robot body. The in-plane LIDAR scanner scans in a circular motion in a plane parallel to the surface of the structure. Using this information, the robot simultaneously constructs a map of its environment and keeps track of its location within it. The accuracy of the simultaneous mapping and localisation mechanism can be improved by inputting data from an accelerometer placed on the robot that can use inertial data to interpolate between 3D scans. Further accuracy improvements can be made by using the BIM model or previous 3D scans to detect features of the environment which have a known position, and then measure the distance of the robot from these known features. This reduces the mapping and localisation mechanism algorithm to a simple localisation mechanism which typically has accuracy on the same order as the accuracy of the LIDAR sensor.

The robot may include a light based positioning system, such as Ultra-Wide Band (UVVB) or infrared laser systems (e.g. HTC VIVE's Lighthouse solution). The light based positioning system uses a set of base-stations at known positions on or near the structure. These base-stations emit light in such a way that by sensing the variations in light intensity a light-sensor can compute its 3D location. By placing sensors on the robot and ensuring line-of-sight between the robot and the base-stations the robot can track its position through space.

Figure 19 shows an exemplary flow diagram for the software the computes the robots location based on a range of sensory information. The diagram is split up into two parts. The part labelled absolute position shows how the global position is calculated using the LIDAR scan data and knowledge of the built environment given by the BIM model and on-site scan data (if available, hence the dashed lines). The part labelled relative position shows how the position of the robot is updated when new scan data from the horizontal scanner that is positioned on top of the main chassis isn't available and updates the position based on detected changes from various other sensors.

Absolute position is calculated by obtaining LIDAR information 19540 about the relative positions of parts of the surrounding environment. Feature detection is then used 19540, using such methods as convolutional neural networks trained to detect BIM objects in LIDAR data or, if as-built data is available 19610, using algorithms that align two different point cloud dataset scans of the same space. Using the known position of these objects and the scan data the robots position can then be computed and updated19560 19570 19580.

In the relative position algorithm, sensor data is taken from the accelerometer 19620, the change in joint positions 19630 and if available vertical scan data 19640 of the brick structure, to compute the change in the robots positionl 9650 19660. Using Kalman filtering these different inputs are combined to produce a more accurate calculation than each sensor input would provide individually. For the case of the scan data we would estimate the number of bricks that the robot had traversed using algorithms trained to detect bricks in point cloud data.

This loop runs continuously while the robot is running 19590.

Structural Unit Positioning In order to accurately place a structural unit, such as a brick, the robot comprises at least one structural unit positioning system which detects the position and orientation of a structural unit as the structural unit is placed on the structure by the robot. Two exemplary structural unit positioning systems are described below.

The robot may include a proximity structural unit positioning system. A proximity structural unit positioning system comprises proximity sensors placed on the gripper which detect the relative distance between the gripper and the brick being placed. Using the position of the gripper, obtained, for example, from the position of the robot and the relative joint positions, and the orientation of the gripper, obtained from a three-axis gyroscope sensor placed on the gripper, the position and orientation of the brick can be computed.

The robot may include a stereoscopic structural unit positioning system. A stereoscopic structural unit positioning system comprises a stereoscopic camera placed on the robot body looking down towards the gripper. Using the global position of the robot, the relative position of the camera and machine learning algorithms trained to measure the position and orientation of a brick from depth-map images, the position and orientation of the brick can be calculated. This has the added benefit of being able to monitor the position and orientation of bricks already placed as the robot manoeuvres over the structure.

Structural Unit Position Correcting The robot comprises at least one structural unit position correcting system for fine tuning the position of a structural unit, such as a brick, when the structural unit is being placed onto a connecting material bed. The structural unit position correcting system comprises at least one structural unit position sensor, which checks the position of the structural unit, and at least one structural unit positioning device, which moves the structural unit. There is a feedback loop between the structural unit positioning device and the sensor to automatically affect the required adjustments to the structural unit's position.

The structural unit positioning device may be the gripper. Alternatively, the structural unit positioning device may be a tapping device, which moves the structural unit vertically and/or horizontally in each corner and/or edge by tapping and/or vibrating on the corresponding part of the structural unit.

The structural unit positioning device may comprise a rotating eccentric mass driven by a small motor, similar to a rotating pile driver or resonant actuator. This motor assembly is mounted on a substructure of the robot which is free to move in one degree of freedom, such that the direction of the force from the structural unit positioning device is in a known direction. The robot may be configured to change the positioning force, as required. The positioning force may be changed, for example, by changing the speed of the motor, and/or changing the resonant frequency of the moving element etc. The robot control software may calculate the required changes to the positioning force from input sensor data and carry out the required taps. This is repeated until the brick is in the desired position, to within a required tolerance.

For example, initially each structural unit positioning device may vibrate individually and the response of the structural unit, i.e. change in position of the brick, be detected. This calibrates the impact on the position of the brick induced by a given force and location of force application. Using this information and the desired alteration of the bricks position, a set of locations and forces can be computed to achieve a desired change in the bricks position. This can be achieved by using, for example, a neural network trained to predict the change in a bricks position given calibration data and a known force and direction of application. The resulting required actions are enacted by each structural unit positioning device and the resulting change in the bricks position is measured using the sensing devices. This cycle repeats until a threshold level of accuracy has been achieved.

Figure 20 is a diagram illustrating an exemplary high level control flow for the robot when placing a brick. The brick is placed initially by the movement of the main chassis of the robot to lower it into place and release the gripper 20670. After the coarse grain placement the robot detects the position of the brick using sensors on the gripper 20680. This gives the x,y,z, yaw, pitch and roll of the brick. This is compared to the position and orientation required by the BIM design 20690. If it is in place to the sufficient degree of accuracy the action is complete. If not the robot calculates the appropriate action tapping action required by the grippers tapping actuators 20700. This action is then performed and the bricks location is again measured and the process repeats 20710. To compute the appropriate and required tapping strengths and positions on the brick machine learning algorithms can be used that make use of the change in the bricks position given previous tapping strengths and positions. By training the algorithm on a wide range of connecting material consistencies and brick types, an effective model can be produced that will adapt to the various conditions encountered in reality.

Figure 31 contains a diagram showing a plan view of an exemplary construction site during the build process. The dashed lines 31320 indicate the brick structure being built. Arrows indicate the direction of travel of the robots. Squares indicate supporting equipment and pathways. Next to the brick structure are the window bypasses 31900, 31901 that indicate that when the brick structure reaches a height at which the robot cannot continually travel along the brick wall the robot can enter the bypass and leave, following the direction of the robot flow, to get around this occlusion. The up-staircase 31820 and down-staircase 31821 represent the staircases as shown in the hardware diagrams, these allow access to all the levels of the brick structure and are accessible from the main supply path 31240. Also connected to the main supply path and accessible is the connecting material supply point 31220, the material supply point i.e. bricks 31220, the wall tie supply point 31221 and the charging station 31960.

Valves and Checkpoints Assembly control involves avoiding collisions, interference and blockages that can slow down the build process as well as being flexible enough to adapt to changing numbers of robots and external input from humans.

The assembly control program sets virtual valves to control the "flow" direction of robots.

Valves are regions of space for which certain directions of movement are prohibited. Valves may be set centrally by the assembly control program. Alternatively, valves may be set by the robots themselves as they traverse the build environment. In this case, robots may treat the valve like a pheromone trail, such that the first robot to approach the valve sets its direction (and the subsequent direction of flow). Valves may be set as a global constraint in that all robots obey the constraints. Alternately, valves may be set as a local constraint in that only certain robots obey the constraints. Valves are useful for circular structures where a continuous "flow" of robots between the build structure and the supply points can be established by setting up effective "one-way streets". This avoids robots having to by-pass each other, which is not always possible for parts of the structure which are narrower than two robot widths. It is also useful in structures where it is possible for robots to by-pass each other, but doing so would slow down the build process. Valves are also useful for enforcing that certain staircases are for climbing up and others for climbing down only.

The assembly control program sets virtual checkpoints to avoid blockages, robot collisions and for general safety procedures. Like valves, checkpoints consist of regions of space. These regions may be defined centrally by the assembly control program. Checkpoints may be set as a global constraint in that all robots obey the constraints of the checkpoint. Alternately, checkpoints may be set as a local constraint in that only certain robots obey the constraints of the checkpoint. When a robot enters a region of space associated with a checkpoint it is required to perform certain checks. For example, the robot may perform safety checks. If the checks fail, the robot may be configured to wait and repeat the checks after a fixed time delay. Alternatively, if the checks fail the robot may be configured to perform some other action. If the checks are passed then the robot may, for example, continue on its way. This might be used to prevent two robots entering a given area of the build, for example. In this case the check might be, is there a robot in a given region of space. If yes, then the robot must wait. If no, the robot may continue on its way. This can be used to prevent collisions between robots and blockages.

Figure 21 is a diagram showing example situations where valves 21720-1 and checkpoints 21730-1 are used to control the flow of robots. These elements combined with built-in collision avoidance are sufficient to ensure the robots do not crash or importantly create blockages that slow the build process. The flow of the robots is indicated by the arrows and the valves and checkpoints are indicated by the squares. The "access to supplies" square 21740 can be thought of as an access staircase that allows the robots access to material supplies and all heights of the brick structure. The upper arrow that connects to the access staircase to the brick structure is the entry point, the lower arrow is the exit point, as indicated by the direction of the arrows. The brick structure 21320 is indicated by the dashed lines.

The upper diagram shows a brick structure with a circular topology. This circular topology allows for the definition of robot flow which is enforced using the valve at the entry point where the robots, upon entering from the supply staircase, are forced to move in an upwards direction. These valves also control the path finding algorithms of the robots. Hence if the robot is finding a path back to the supply staircase, its only viable route is to walk all the way around the brick structure to the exit point. This ensures that no robot turns back at some other part of the brick structure where there are no valves. This flow setup enables multiple robots to walk around the structure without blocking each other's way. The only region where this unidirectional flow is not possible is in the region between the entry and exit points. To lay bricks in this region and then access the exit point the robot must walk back in the direction opposite to the flow in other regions. This is not essential but saves the robot walking around the entire structure twice. To prevent blockages in this region a checkpoint has been added. Robots wait here if there is another robot in the region between the entry and exit points.

The lower diagram shows the flow control for a single brick wall structure. This structure has a linear topology so no circular current or flow can be established. Robots must walk back and forth along this structure and potentially block each other's way. To avoid this, a checkpoint is located at the entry point to the brick structure from the supply staircase. Robots wait at this checkpoint if there is another robot on the brick structure.

Additional Hardware Windows, doors and other openings in a brick wall occlude the path of the robots, preventing them from accessing all parts of the brick wall/structure. In order to overcome this, the automated assembly system may include at least one bypass system that, when placed next to the openings in the brick wall, enable a continuous path from one side of the opening to the other. The bypasses comprise a support structure e.g. columns that uphold level surfaces that form a path out away from the plane of the wall and round to the other side of the opening. In a preferred embodiment, these level surfaces occur every 3 courses, ensuring that a path is accessible at all intermediate heights of the brick structure using an up, down or level step by the robot. To ensure enough head clearance for the robot to be able to move along the bypass the level surfaces alternate in the extent of their protrusion away from the plane of the brick structure. Thus, the total head room is twice the height of 3 courses.

Figure 30 shows an exemplary perspective view of a design for a bypass structure, here shown enabling the robot to bypass a window gap in the brick wall structure 30320 shown behind. The bypass consists of a main supporting frame 30910, shown as the thick cylindrical poles or supports. This frame upholds level surfaces traversable by the robot that enables continuous access from one side of the window to the other. The robot is illustrated manoeuvring on one of the level surfaces 30920 at the same height as the current brick course. The level surfaces occur every three brick courses, ensuring that a path is accessible at all intermediate heights of the brick structure using an up, down or level step by the robot. Every alternate level surface illustrated in the diagram provides a path that protrudes further from the brick wall, going around the level surfaces closer to the brick wall.

The robot illustrated is shown on such a path. This ensures enough head clearance for the robot to be able to move along the bypass. Thus, the total head room is twice the height of 3 courses.

Figure 29 shows a side on view of an exemplary bypass structure. On the far left of the figure, moving from left to right, is a representation of the inner-leaf 29940, insulation 29950 and the brick wall structure 29330 respectively. On the far right we see extending vertically the support frame 29910, holding up five attached structures. These attached structures are the level surfaces eg 29920 that provide pathways for the robot to traverse from one side of the gap in the brick wall (in this case a window) to the other. Each of the five attachments to the supporting frame holds up two such pathways, one at the same height as the attachment point 29920 (such as the one the robot is illustrated standing on in the middle) and another three brick courses higher 29930 giving a total of ten level-surface pathways. It can be seen that the level-surface pathways occur every three brick courses, ensuring that a path is accessible at all intermediate heights of the brick structure using an up, down or level step by the robot. The alternating design of the level surfaces, such that every odd pathway extends around and out away from the brick surface gives a head clearance twice the height of three brick courses, sufficient for the robot.

Figure 28 shows a perspective view of an exemplary bypass structure on its own. It shows the supporting frame 28910 at the back of the image, with the attached level-surface 28920- 4 28930-4 pathways for the robot in the foreground. See descriptions of Figure 9 and Figure for more information.

The automated assembly system may include at least one staircase system. A staircase system can be used to enable the robots to climb up and down between different levels of the structure. In a preferred embodiment, a staircase system comprises two spiral staircases that are supported by vertical support columns. In this way, one robot may climb from the bottom of the structure to the top of the structure at the same time as a second robot climbs from the top of the structure to the bottom of the structure. Each spiral staircase provides structure access points at which the robot may access the structure. If a robot is only able to climb up or down one course at a time, the access points may be provided at heights corresponding to every three courses of the structure to be built. This ensures the robot may access the structure at all intermediate stages of the build. To see this, take an example of a group level access point at course 0. The robot can trivially access course 0 by taking a level step from the ground level access point. Similarly, the robot can access course 1 of the brick structure by taking a step up from the ground level access point. Course 2 can be accessed by the robot by taking a step from the ground level access point on to course 1 and then placing a brick on top of course 1, creating course 2. As long as course 2 is laid near the access point at course 3, the robot can then continue to access course 2 of the brick structure from the access point at course 3 by taking a step down. The spiral staircase design gives a total headroom for the robot of twice the height of three brick courses. The staircase may also comprise a pathway that extends away from the structure to enable access to a supply point, such as a brick pallet of cut or full size bricks. The staircase system may be powered such that the robots can charge the robot batteries as the robots use the staircase system.

The staircase provides access every three courses because the robot, from any given level can access three levels. That is, the same level, the level one below and the level one above. The robot can also lay bricks on any of these levels. If we take the case where the robot is accessing the brick structure from a staircase at course 3, the robot can step up onto the structure and thus be walking on course 4. Once on this level the robot can lay bricks, thus laying the bricks associated with course 5. Course 5 is accessible via a step down from the staircase access point at course 6, three courses above the current access level. Thus if the robot lays the bricks in front of this higher access point (at the height of course 6) then it can start accessing the brick structure at this level.

Figure 25 shows an exemplary staircase 25820 that enables the robot 25000 to access different heights of the brick structure. The staircase consists of a supporting frame that upholds level surfaces 25840 forming the stepping and walking points for the robot. These level surfaces form two spiral staircases, one on the left and one on the right, enabling two robots to walk up and down simultaneously without having to pass each other. The difference in height of the level surfaces that form the steps is equivalent to the height of a brick. The supporting frame consists of outer vertical support structure 25850 25851 and an inner frame 25860 that holds up the level surfaces forming the staircases. The outer vertical support structure consists of two vertical poles 25870 25871 on either side held together with five cross attachments 25880-4. An extra angled supported pole is included for balance 25872 25874 that attaches at the top. The inner frame is detachable from the outer supporting frame enabling the height of the staircase to be changed depending on the requirements. The inner frame is hinged at the points where the parts that uphold the level surfaces meet the parts that connect to the outer supporting frame. These are illustrated by circular protrusions in the diagram. This enables the inner staircase frame to be folded-up and packed efficiently for distribution and transport. The diagram shows the robot placed upon the ground level surface of the staircase that protrudes away from the brick wall (not shown). This level-surface provides access to, for example, brick supplies and charging stations etc. Access to the wall is provided every three brick courses to enable the robot to step on to the wall at all intermediate heights during the build process.

Figure 27 illustrates the design of the staircase 27820 shown in Figure 25 positioned in front of a brick wall 27320. A robot is shown positioned on the level surface that provides access to the brick wall at its current height with a step-down. The outer supporting frame 27850 27851 can be seen on either side of the staircase. The inner-frame 27860 supporting the level surfaces 27850 that the robot manoeuvres on is shown, with the circular hinges visible at each level of the staircase.

Figure 26 shows a side-on view of the staircase design. It includes a brick wall 26320 structure on the far right, accessible via the staircase 26850, and a brick supply pallet on the far left 26330. The robot 26000 is positioned on the ground floor pathway 26240 of the staircase that provides access to the brick pallet on the far left. Here we can see clearly that the access points from the staircase to the brick structure occur every three brick courses and the alignment of the steps with the brick structure.

Alternative embodiments of the staircase may be designed to allow the structure to be assembled by a robot. That is, each stepped element may be a modular component. The robot would be able to climb the staircase thus assembled and add to the top of the staircase. This process could also be performed in reverse at the end of the build process.

To speed up the vertical movement of robots up to higher levels, the automated assembly system may include at least one lift system. An example lift system uses compartments or platforms mounted on a vertical rail, controlled with a motor and balanced by a counterweight. The lift system includes a control program which communicates with the robots, such that robots wait in designated 'lobby' areas. This prevents robots from blocking access to the lift, or to other areas in the automated assembly system. Further, the lift and/or designated lobby areas may provide charging facilities for the robots 20.

Figure 22 shows a perspective view of an exemplary robot vertical transportation device or lift 22750. This consists of a primary vertical structure, a linear rail mounted vertically 22760 on which moves a platform or lift car 22770. This is driven by a motor, and balanced with a counterweight 22780. Alongside the platform, there are stationary platforms 22790-4 attached to the primary structure, positioned to allow movement by the robot between the stationary platforms and the moving lift platform. The whole unit is positioned such that it is possible for the robot to move between the structure it is building and the rest of the work area, and the moving lift platform and the stationary platforms. The platforms are arranged vertically such that the step height between each platform and each brick level of the target structure is within the robot's capabilities, and also that there is sufficient clearance height between stationary platforms for the robot to stand. Here, the platforms are arranged each third course of bricks, assuming the robot can step up one brick or down one brick height. The platforms are also alternating each side to allow the clearance height of approximately six bricks. The robot may interact with the lift's control software using a variety of methods such as pressure, proximity, visual sensors etc, or may communicate directly via wireless communication.

Figure 23 shows a side elevation view of an example robot vertical transportation device 23760. It shows an example built structure 23320 composed of bricks, with the stationary platforms 23790-3 attached to the lift primary structure arranged so as to interface with the built structure in intervals to allow the robot to move between the structure and the platforms.

It shows the moving lift platform mounted to the primary vertical structure, supported by a cable element. This cable is attached to a counterweight 23780, via a pulley which is driven by a motor 23800.

Figure 24 shows an exemplary sequence of perspective views of an example robot vertical transportation device. It shows the purpose of the stationary platforms being used as waiting areas by different robots. This allows the lift to be used effectively by multiple robots. The first diagram shows one robot 24000 entering the lift at the lower level, aiming to deliver its brick to the built structure. The second diagram shows another robot 24001 entering onto one of the stationary platforms on its respective level. It has been informed by communication with the lift or general robot network that the lift is currently at the bottom, so it must wait for the robot currently in the lift to exit before it enters. The third diagram shows the lift having been raised to the current working level. Because the returning robot 24000 is waiting on the stationary platform, the lift exit is clear for the outcoming robot 24001 to leave the lift and carry its brick to the built structure. The fourth diagram shows the returning robot having side-stepped onto the lift platform in order to be taken down to the lower level.

Supply Point The automated assembly system may comprise a structural unit supply point. The structural unit supply point provides a supply of structural units, such as bricks. The structural unit supply point should be accessible to the robots from the structure. The structural unit supply point may enable the robots to obtain bricks one or many at a time. As well as being accessible to the robots, the structural unit supply point needs to be easily accessed by the structural unit supply delivery mechanism e.g. truck, crane, or forklift.

A collection of structural units may be provided as a plurality of layers of structural units, referred to herein as pallets. The pallet arrangements can be achieved using full-bricks or cut-bricks of differing shapes and sizes. The bricks may stacked in alternating orientations to allow the bricks to interlocking through gravity-induced friction. The layers may also be separated by a sheet material such as paper, card or plastic sheet. The pallet size and arrangement can vary. The larger the pallet, the greater the autonomy of the overall system -e.g. a pallet could be pre-loaded in the factory with all the bricks required for a given structure (e.g. a whole house), removing the need to resupply the automated assembly system with additional pallets during the assembly process.

The automated assembly system may comprise a staircase to allow the robots to access the upper levels of a pallet. This staircase may be provided using the structural units themselves: a plurality of the layers of structural units in a pallet will each comprise less structural units than the layer on which it rests. In this way, the pallet could be considered an 'incomplete' pallet, as it contains less structural units than would be included in a pallet by legacy structural unit suppliers. Alternatively, the automated assembly system may comprise an additional staircase system that is set down beside the pallet to allow the robot to climb to the top of a complete pallet. This ensures that the system can be used with structural units supplied by legacy structural unit suppliers.

A dedicated container or special large pallet can be used to transport the structural units, such as bricks, from the factory to the site. The dedicated container may include an inverse staircase feature. The inverse staircase feature 'completes' the pallet. The inverse staircase feature extends into the space in a layer of structural units for which no structural unit is present, thereby preventing the structural units from moving around in the container when the structural units are transported.

Figure 14 shows a perspective view of an exemplary pallet of bricks 14330 which have been specially prepared to allow robot access. The bricks are arranged into a standard pattern. These are placed on a pallet or structural member to allow easier transportation. Bricks have been left out or removed from this arrangement to allow a natural staircase in once brick increments leading up one side of the pile, to allow the robot to reach all the levels. Then a filler piece 14340 consisting of a shaped solid body, for example made from foam, timber, or metal is inserted to fill this void. The filler piece helps keep the bricks in their proper positions during transportation, and makes the whole pile in a regular box volume.

Figure 15 shows the above pallet 15330 after it has been packed. The filler piece 15340 has been put in place, and the entire ensemble has been strapped together with bands 15350-1 to secure it in place, and may also be covered with a plastic film to keep it protected from moisture. The addition of the filler piece keeps the otherwise unstable form secure during transport and allows the stacks to be tessellated with one another neatly.

Figure 16 shows an exemplary large container for the transportation and storage of a large volume of bricks. The unit consists of a structural platform, with removable or hinged side pieces 16380-1. The unit has integrated features to allow easy lifting 16360-3 by an external device such as a crane. It also has a built in or modular attachment staircase 16390 to allow robots to climb up to the top of the brick pile and retrieve bricks. The end of the staircase is connected to the rest of the build area with dedicated pathway 16370, to assist the robots to gain access to the pile. This is shown here integrated into one of the side panels which has been folded down, but may also be a modular attachment.

The automated assembly system may comprise a pallet lift, on which pallets can be placed. The lift is positioned such that it may raise the pallets to a position wherein the top of the pallet is accessible to the robot from the top of the structure. The lift need not be next to the structure. There may be a raised pathway that enables robots to manoeuvre from the structure to the point at which the lift raises the pallets. In this case the staircase design of the structural unit pallet is not required as the top of the pallet is directly accessible from the structure.

The pallets can be placed at the base of the staircase design described above which is easily accessible to delivery trucks and cranes. In this case, the staircase design of the structural unit pallet is not necessarily required, as the top of the pallet may be directly accessible from access points on the staircase.

Figure 17 shows a perspective view of a storage area and integrated staircase for bricks on site. The system consists of a flat platform, an integrated or detachable staircase unit 17390, an integrated or detachable robot walkable pathway 17370 leading to the rest of the build area, and a pack of bricks 17330. The storage area would be set up in position such that the robots in the system are aware of its location. Bricks can be added in whole packs at a time -they are simple placed on the platform. The integrated staircase allows the robots to climb up and access the high levels. By removing bricks in a deliberate order e.g. picking those furthest away on each level, and working down level by level, the pack of bricks always remains traversable during the build. The robots can monitor the level of the stack and given notification when more bricks are required.

The automated assembly system may comprise an additional pathway on which the robots may manoeuvre. The additional pathway may, for example, be placed between the structure and a supply point, or between the structure and a charging point. In one example, a pathway is placed between the structure and a pallet, such that the pallet may be placed away from the structure in a location that is easily accessible for pallet delivery systems. The additional pathway may consist of modular pieces which can be fitted together to create a pathway of any length and to any direction. The pathway may include turns and/or stairs to allow the pathway to reach any destination, and to avoid any obstacles. The pathway may include bifurcations to provide paths to different parts of the automated assembly system. For example, a pathway from a first point on the structure may bifurcate into four different paths -the first branch leading to a material supply point; the second branch leading to a charging station; the third branch leading to a maintenance point; and the fourth branch leading to a second point on the structure, different from the first point on the structure. The surface of the pathway may be similar to the top surface of the brick type being used for construction. Alternatively, the surface of the pathway may have a standard surface, designed to be manoeuvred by robots using different feet.

The automated assembly system may comprise a reshaping point, accessible to the robot. This reshaping point may be used to reshape structural units as required. For example, a reshaping point could be placed near a brick supply point, and used to cut bricks as required. The reshaping point may comprise a powered circular or band saw fixed in place.

The robot may connect the structural unit to the reshaping point, the reshaping point then reshapes the structural unit, as required, and returns the brick to the robot. Alternatively, the reshaping point may be configured such that the robot may hold bricks up to a blade to cut them to specified shapes (e.g. half bricks, angled bricks etc.). This would mean more onsite equipment is required, but would allow more brick customisation at the point of use, and hence less advance planning in the factory.

The automated assembly system may comprise a connecting material supply point. The robot comprises a connecting material storage and delivery system. The connecting material storage and delivery system comprises a connecting material container for holding the connecting material used to lay a brick. In order for the robot to refill its connecting material storage and delivery system, the automated assembly system may comprise a connecting material supply point. The connecting material supply point comprises a specially designed connecting material silo that interfaces with the robot's connecting material storage and delivery system. The connecting material silo is positioned above a walkable pathway for the robot. The connecting material silo comprises a nozzle that hangs down over the pathway at a height accessible to the robot. The robot is able to walk beneath the connecting material silo nozzle and lift its body. The connecting material silo nozzle enters an opening in the top of the robots connecting material storage and delivery system, for example the top of the connecting material container. The connecting material storage and delivery system opens as the robot raises the main beam. Surrounding the opening in the connecting material storage and delivery system is a seal that ensures an air tight fit between the robot container and the silo nozzle. The seal may be made of a flexible material such as rubber. A sensor such as a proximity sensor on the side of the silo nozzle detects the presence of the robot and triggers the connecting material to flow from the silo nozzle in to the robots canister. The connecting material silo detects when the robots canister is full using, for example, a mechanism similar to that found in petrol pumps. A small tube in the silo nozzle is connected to a suction device such as a venturi. Whilst the robot canister is being filled air flows through this small tube. Once the canister is full the air tube is blocked by the connecting material and the change in suction is detected triggering the connecting material silo to close its nozzle. Another approach is to use pressure sensors in the feet of the robot to detect the change in weight of the robot. Once the weight of the robot corresponds to the weight of the robot when fully laden with connecting material, the robot can trigger the connecting material silo to close the nozzle.

Figure 6 shows a perspective view of an exemplary connecting material storage unit 6220, robot unit 6000 and custom dispenser apparatus 6230 which interfaces between the two.

The connecting material storage unit is a large hollow container, mounted above the ground. It may have an internal auger screw or paddle to assist with connecting material mixing, and integrated sensors and control program to monitor the quantity, moisture or chemical content of the connecting material. The storage unit is positioned above a walkable pathway 6240 for the robot, such that the robot can walk along and stop underneath the storage unit. The unit has an opening through which connecting material can flow from the unit into the robot. To do this, it passes through a flow valve and nozzle, which is connected to a pressure or distance sensor positioned to interface with the robot. The nozzle and the opening of the top of the robot are designed in conjunction, such that the nozzle will fit snuggle into the robot's opening 1080. The unit's height is such that the robot can lower itself on its legs 1010-1, position itself underneath the unit's nozzle, and then raise itself up on its legs, thus inserting the nozzle into its opening. This would automatically begin connecting material flow, through contact or interface with a sensor on the storage unit. Connecting material flow may be measured through a variety of ways such as flow sensor in the nozzle 5200, depth sensor in the robot's container, or strain gauges built into the robot's legs detecting the change in the robot's mass as connecting material is delivered. Once sufficient connecting material has been transferred, the robot can lower itself on its legs, thus stopping the connecting material flow and retracting the nozzle from the opening, and then walk onwards to lay a brick.

Figure 7 shows perspective views of an exemplary robot refilling itself with connecting material from an exemplary storage unit 7220. The first view shows the robot lowering itself on its legs and positioning itself underneath the storage unit's nozzle 7230. The second view shows the robot having raised itself up on its legs, thereby inserting the storage unit's nozzle 7230 into its opening. The opening's cap 4120-1 has been pushed apart by the nozzle as it has been inserted. The sensor on the storage unit has made contact with the robot's body, thereby initiating the connecting material to start flowing. (See above for details).

The automated assembly system may comprise a wall-tie supply point. Wall-ties are an integral part of brick outer-leaf structures. They ensure that the wall is structurally safe by securing it to the inner-leaf. A wall-tie supply point comprises a wall tie dispensing system. The wall tie dispensing system comprises a stack of wall-ties contained in a cartridge like container. At the base of the cartridge like container is an opening and a mechanism (active or passive) that pushes out the wall-tie at the bottom of the stack. This enables a robot with the appropriate gripper to take the wall-tie from the cartridge. When the wall-tie is removed the stack, the wall-ties fall under gravity to fill the place of the removed one.

Charging Point The automated assembly system may comprise a robot charging point. A robot charging point allows the robots to charge their batteries autonomously. The robot charging point may be powered either by the mains or a portable generator, e.g. petrol or solar power. The robot charging point may be an area with a number of docking stations and respective pathways to enable robot access. Alternatively, the robot charging point may include only one docking station. The docking stations could allow access to the rear of the robots main beam. The docking station comprises a conductive contact that connects with the charging ports on the rear of the robot's body. In this way, the robot is able to walk directly into the docking station to connect with the conductive contact. Upon detection of the robot the charging station initiates the charging process. The robot charging point could also use wireless charging through induction.

For faster power replenishment, robot charging points could also allow batteries to be exchanged. In this case, the battery would be a removable part of the robot, for example, a removable part of the robot body (e.g. snap fit or release catch). Batteries could be deposited onto vacant charging racks or slots, and re-charged ones picked up in place.

A robot charging point may be included in other parts of the automated assembly system.

For example, power may be provided via a staircase, a lift, and/or a pathway. A robot may include a connection on its feet to connect with the charging point on the staircase, lift, and/or pathway. The charging point may connect with the robot at a plurality of points on a part of the automated assembly system. By including connections on both feet, the robot will always be in contact with the staircase, lift, and/or pathway whilst it manoeuvres around them. Therefore, the robot will continually charge. Alternatively, the robot may charge via induction.

Communication The robots are configured to communicate with communicating parts of the automated assembly system in order to share information about their status and their actions, and receive information from the other communicating parts of the automated assembly system.

The other communicating parts of the automated assembly system may be other robots, or any other part of the automated assembly system which communicates with the robots, such as a central computer.

Communications in the automated assembly system may be wireless. Wireless communications may be based on Bluetooth, W-Fi and/or broadband cellular network technology, for example. Communication may also, or alternatively, be achieved through a physical connection to a part of the automated assembly system. For example, a staircase or pathway may include a connection which connects a robot to a communication channel as the robot manoeuvres on the staircase or pathway. Similarly, a charging point may include a connection which connects a robot to a communication channel. Communication may be achieved via a connection used to transmit power to the robot.

The network structure with which communicating parts of the automated assembly system communicate with other communicating parts of the automated assembly system may comprise any network topology, some examples of which are described below.

A communicating part of the automated assembly system may be configured to communicate with other communicating parts of the automated assembly system directly.

Each communicating part of the automated assembly system in the system may be able to communicate with each other communicating part of the automated assembly system directly. Alternatively, a communicating part of the automated assembly system may only be able to communicate directly with a subset of the communicating parts of the automated assembly system. The subset of communicating parts of the automated assembly system with which a particular communicating part of the automated assembly system may communicate may depend on the position or type of the communicating parts of the automated assembly system -for example, whether it is another robot, or a charging station, and where the robot or charging station is in relation to the part of the automated assembly system that wishes to communicate. For example, robots may communicate with other robots using a distributed full-mesh network, where each robot communicates directly with each other robot.

A communicating part of the automated assembly system may be able to communicate with other communicating parts of the automated assembly system indirectly via the communicating parts of the automated assembly system with which it may communicate directly -for example, through a central computer. This creates a centralised network where communications between two communicating parts of the automated assembly system pass through a third communicating part of the automated assembly system.

The communicating parts of the automated assembly system may be configured to communicate via mixed direct and indirect communication. This approach uses some direct communication between communicating parts of the automated assembly system whilst also using indirect communication between communicating parts of the automated assembly system -for example, using direct communication between robots and indirect communication between robots other communicating parts of the automated assembly system (such as a material supply point that is configured to communicate with the robots) via a central computer.

Figure 33 contains three diagrams of the possible network connections between the robots and a central controller. The arrows indicate information pathways transmitted either by Bluetooth, Wi-Fi or cellular network. The upper diagram shows a fully connected mesh network where each robot 33000-4 is connected with every other robot. The middle diagram shows a centralised start network where all robots are connected to the central controller 33810 and can only communicate with each other through this central station. The final diagram shows a fully connected mesh network where all robots can communicate with each other and a central controller. This last option enables the advantages of direct communication with the option of being able to control all robots simultaneously, for example allowing for a system wide shutdown in an emergency.

As will be apparent to the skilled person, the exemplary features described herein may be combined with other robots, and other systems. For example, a connecting material delivery system may be included on another robot.

The following is a list of aspects of the invention: 1. A robot for use in an assembly system in which a structure is assembled out of a plurality of structural units, wherein the robot is configured to manoeuvre around the structure.

2. A robot according to aspect 1, wherein the robot is configured to manoeuvre around the structure by traversing and/or climbing the structure.

3. A robot according to any preceding aspect, wherein the robot comprises: a robot body; and a first foot coupled to the robot body and a second foot coupled to the robot body, wherein the first and second feet are configured to move relative to the robot body and each other.

4. A robot according to aspect 3, wherein the first and second feet are coupled to opposite sides of the robot body.

5. A robot according to aspect 4, wherein the first and second feet are coupled to the opposite sides of the robot body by legs, wherein the leg of each foot is configured to move in a respective movement plane, horizontally and vertically relative to the robot body, wherein the sole of each foot extends horizontally towards the movement plane of the opposite leg.

6. A robot according to aspect 5, wherein at least a portion of the legs are spaced apart to allow a structural unit to pass between them as the legs pass each other.

7. A robot according to any preceding aspect, wherein the robot further comprises an effector coupled to the robot body.

8. A robot according to aspect 7, wherein the effector is configured to releasably couple to a structural unit.

9. A robot according to aspect 7 or 8, wherein the robot is configured to arrange the robot body and effector so that the robot centre of mass is over one foot, then lift the other foot, then move the raised foot past the stationary foot whilst the robot body and/or main effector are moved in the opposite direction, then place the raised foot down.

10. A robot according to aspect 9, wherein the robot is configured to arrange the robot body so that the centre of mass is over the other foot.

11. A robot according to any preceding aspect, wherein the robot comprises a structural unit positioning system.

12. A robot according to aspect 11, wherein the structural unit positioning system comprises proximity sensors placed on the effector.

13. A robot according to aspect 11 or 12, wherein the structural unit positioning system comprises a stereoscopic camera placed on the robot body.

14. A robot according to any preceding aspect, wherein the robot comprises a structural unit position correcting system, wherein the structural unit position correcting system comprises at least one structural unit position sensor and at least one structural unit positioning device.

15. A robot according to aspect 14, wherein the structural unit position device is a tapping device.

16. A robot according to aspect 14, wherein the structural unit position correcting system is configured to change the position of the structural unit by tapping and/or vibrating on the structural unit.

17. A robot according to any of aspects 14 to 16, wherein the at least one structural unit position sensor, is configured to check the position of the structural unit, and the at least one structural unit positioning device changes the position of the structural unit.

18. A robot according to any preceding aspect, wherein the structural unit is a masonry brick.

19. A robot according to any preceding aspect, further comprising the structural unit, wherein the structural unit is preferably a masonry brick.

20. A robot according to any preceding aspect, wherein the robot is configured to pick up bricks and connecting material from a material supply point, manoeuvre around the structure to a position on the structure, apply the connecting material, place the brick, and then climb back to a material supply point or a charging point.

21. A robot according to any preceding aspect, wherein the robot comprises a connecting material storage and delivery system.

22. A robot according to aspect 21, wherein the connecting material storage and delivery system comprises a connecting material container connected to a nozzle.

23. A robot according to aspect 22, wherein the nozzle is a detachable element that is detachable from the remainder of the connecting material storage and delivery system.

24. A robot according to aspect 22 or aspect 23, wherein the nozzle is angled to allow connecting material exiting the connecting material storage and delivery system to be disposed by the robot on both horizontal and vertical surfaces.

25. A robot according to aspect 22, aspect 23, or aspect 24, wherein the nozzle has a cover which covers the opening of the nozzle.

26. A robot according to any of aspects 22 to 25, wherein the connecting material storage and delivery system is configured to build up pressure behind the nozzle cover before releasing the connecting material.

27. A robot according to any of aspects 21 to 26 wherein the connecting material storage and delivery system is configured to apply the connecting material in a shape that is taller and narrower than the intended volume of the connecting material when it has a structural unit in place on top of it.

28. A robot according to any of aspects 22 to 27, wherein the connecting material container is covered by a moveable cap.

29. A robot according to aspect 28, wherein the connecting material storage and supply system is configured to interface with a connecting material supply point, wherein the moveable cap is configured to open when the connecting material supply point is connected to the connecting material container.

30. A robot according to aspect 29, wherein the robot is configured to position the moveable cap underneath the connecting material supply point and then raise the robot body such that the moveable cap is opened, and the connecting material container is connected to the connecting material supply point.

31. A robot according to any of aspects 22 to 30, wherein the nozzle is configured to rotate around a vertical axis.

32. A robot according to any of aspects 22 to 31, wherein the horizontal position of the nozzle is fixed relative to the robot body such that movement of the robot body in a horizontal direction moves the nozzle in a horizontal direction.

33. A robot according to any of aspects 22 to 32, wherein the nozzle is configured to move vertically relative to the robot body.

34. A robot according to any of aspects 22 to 33, wherein the connecting material storage and delivery system comprises a reservoir for storing a connecting material additive for adding to the connecting material.

35. A robot according to aspect 34, wherein the additive is water, wherein the connecting material storage and delivery system comprises at least one sensor to sense the moisture content of the connecting material in the connecting material storage and delivery system, wherein the connecting material storage and delivery system is configured to control the moisture content of the connecting material in the connecting material storage and delivery system by adding water from the reservoir.

36. A robot according to aspect 34, wherein the additive is added to the connecting material as it is expelled from the nozzle, wherein the additive increases the rate at which the connecting material sets.

37. A robot according to any of aspects 22 to 36, wherein the nozzle comprises a device for finishing the connecting material after a structural unit has been laid.

38. A robot according to aspect 37, wherein the device is a pointing element for pointing the connecting material after a structural unit has been laid.

39. A robot according to any preceding aspect, wherein the robot comprises a wall tie applicator.

40. A robot according to aspect 38, wherein the robot comprises a connecting material storage and delivery system, wherein the wall tie applicator comprises a wall tie gripper mounted on the rear of the connecting material nozzle.

41. A robot according to any preceding aspect, wherein the robot comprises a robot balancing system.

42. A robot according to aspect 41, wherein the robot balancing system uses a measurement of at least one of pressure, strain, and changes in angular momentum to balance the robot.

43. A robot according to any preceding aspect, wherein the robot comprises a collision avoidance system.

44. A robot according to any preceding aspect, wherein the robot comprises a robot control program, wherein the robot control program is at least partially controlled by an assembly control program.

45. A robot according to any preceding aspect, wherein the robot comprises at least one positioning system for determining the robot's position on the structure.

46. A robot according to aspect 45, wherein the positioning system determines the robot's position on the structure by calculating a change in the robot's position based upon the movement of the robot.

47. A robot according to aspect 45 or aspect 46, wherein the positioning system comprises a sensor for identifying discrete structural units around the robot's feet, wherein the positioning system is configured to use the metrology of the structural units as discrete objects to determine the robot's position on the structure.

48. A robot according to any of aspects 45 to 47, wherein the positioning system comprises a sensor for determining a model of the robot's local environment, wherein the robot is configured to compare the model of the robot's local environment with a model of the structure to determine the robot's position on the structure.

49. A robot according to any of aspects 45 to 48, wherein the positioning system comprises a sensor configured to detect light emitted from stations at known positions on or near the structure, wherein the positioning system computes its location by sensing variations in light intensity.

50. A robot according toany preceding aspect, wherein the robot is configured to determine a map of an area of the structure greater than the locality of the robot by determining a map of the structure in the locality of the robot at each of a plurality of positions on the structure.

51. A robot according to aspect 50, wherein the robot is configured to compare the determined map with a model of the structure in order to determine the robot's position on the structure.

52. A robot according to aspect 1, wherein the robot further comprises an accelerometer that uses inertial data to interpolate between the maps determined at the plurality of positions.

53. A robot according to any preceding aspect, wherein the robot is configured to obey valve constraints which determine in which directions the robot may travel on a path.

54. A robot according to any preceding aspect, wherein the robot is configured to perform checks at checkpoint positions on the structure.

55. A robot according to any preceding aspect, wherein the robot comprises a power source, wherein the robot is configured to charge the power source as it manoeuvres around additional hardware in the assembly system 56. A robot according to any preceding aspect, wherein the robot is configured to communicate with other robot's using a distributed full-mesh network.

57. A system for assembling a structure out of plurality of structural units, wherein the system comprises at least one robot, preferably a plurality of robots, according to any preceding aspect.

58. A system for assembling a structure out of a plurality of structural units, wherein the system comprises at least one robot, preferably a plurality of robots, configured to manoeuvre around the structure.

59. A container for transporting a collection of structural units arranged in layers, wherein at least one layer of the collection of structural units includes fewer structural units than a layer below it, wherein preferably the container comprises an inverse staircase feature which prevents the structural units in a layer with fewer structural units than a layer below it from moving relative to the layer below it.

60. The robot according to any preceding aspect, wherein the robot further comprises an effector configured to releasably couple to a structural unit.

61. A method of organising at least one robot to assemble a structure out of a plurality of structural units, wherein the method comprises: receiving a model of a structure; determining a structural unit layout model of the structure; and determining a sequence of assembly stages such that the structure is accessible to the at least one robot at all assembly stages.

62. A method according to aspect 61, wherein the received model of the structure is a 3-dimensional model of the structure.

63. A method according to aspect 61 or aspect 62, wherein the structural unit layout model of the structure determines the position of each of a plurality of structural units in the structure.

64. A method according to aspect 63, wherein determining the sequence of assembly stages includes determining the order for laying the structural units.

65. A method according to any of aspects 60 to 64, wherein the method further comprises determining a flow control for a plurality of robots.

66. A method according to aspect 64, wherein determining the flow control comprises determining the position of a valve on a path on which a robot manoeuvres between points in the automatic assembly system, wherein the valve determines the direction of movement the robot may take.

67. A method of organising at least one robot in an assembly system, wherein the method comprises: setting a valve on a path on which a robot manoeuvres between points in the automatic assembly system, wherein the valve determines the direction of movement the robot may take; communicating the position of the valve to the robot.

68. A method according to aspect 67, wherein the method further comprises determining the path on which the robot manoeuvres between points in the automatic assembly system.

69. A method according to aspect 67 or aspect 68, wherein the valve is set at a point on the path, and the valve determines the directions in which the robot may move from that point.

70. A method according to any of aspects 67 to 69, wherein the valve is set along a section of the path, and the valve determines the directions in which the robot may move along that section.

71. A method according to any of aspects 66 to 70, wherein the valve is set by a robot as it manoeuvres around the assembly system.

71. A method of organising at least one robot in an assembly system, wherein the method comprises: setting a checkpoint on a path on which a robot manoeuvres between points in the automatic assembly system, wherein the robot performs a check when it reaches the checkpoint; communicating the position of the checkpoint to a robot.

72. A method according to aspect 68, wherein the method further comprises determining the position on which the robot manoeuvres between points in the automatic assembly system.

73. An assembly system for assembling a structure out of a plurality of structural units, wherein the system comprises: a staircase system; at least one robot configured to access or exit the structure via the staircase system.

74. A staircase system for use with an assembly system comprising at least one robot.

75. A staircase system according to aspect 73 or 74, wherein the staircase comprises a plurality of steps between a plurality of access levels, wherein the access levels are at heights corresponding to every three steps.

76. An assembly system for assembling a structure out of a plurality of structural units, wherein the system comprises: a bypass system; at least one robot configured to manoeuvre between points on the structure via the bypass system.

77. A bypass system for use with an assembly system comprising at least one robot.

Claims

CLAIMS1. A robot for use in an assembly system in which a structure is assembled out of a plurality of structural units, wherein the robot is configured to manoeuvre around the structure.
2. A robot according to claim 1, wherein the robot is configured to manoeuvre around the structure by traversing and/or climbing the structure.
3. A robot according to any preceding claim, wherein the robot comprises: a robot body; and a first foot coupled to the robot body and a second foot coupled to the robot body, wherein the first and second feet are configured to move relative to the robot body and each other.
4. A robot according to claim 3, wherein the first and second feet are coupled to opposite sides of the robot body.
5. A robot according to claim 3 or 4, wherein the robot comprises a main effector, wherein the robot is configured to arrange the robot body and main effector so that the robot centre of mass is over one foot, then lift the other foot, then move the raised foot past the stationary foot whilst the robot body and/or main effector are moved in the opposite direction, then place the raised foot down.
6. A robot according to any preceding claim, wherein the robot comprises a structural unit positioning system.
7. A robot according to any preceding claim, wherein the robot comprises a structural unit position correcting system, wherein the structural unit position correcting system comprises at least one structural unit position sensor and at least one structural unit positioning device.
8. A robot according to any preceding claim, wherein the robot comprises a connecting material storage and delivery system.
9. A robot according to claim 8, wherein the connecting material storage and delivery system comprises a connecting material container connected to a nozzle.
10. A robot according to any preceding claim, wherein the robot comprises a wall tie applicator.
11. A robot according to any preceding claim, wherein the robot comprises a robot balancing system.
12. A robot according to claim 11, wherein the robot balancing system uses a measurement of at least one of pressure, strain, and changes in angular momentum to balance the robot.
13. A robot according to any preceding claim, wherein the robot comprises a collision avoidance system.
14. A robot according to any preceding claim, wherein the robot comprises at least one positioning system for determining the robot's position on the structure.
15. A robot according to any of claims 14, wherein the positioning system comprises a sensor for determining a model of the robot's local environment, wherein the robot is configured to compare the model of the robot's local environment with a model of the structure to determine the robot's position on the structure.
16. A robot according to any preceding claim, wherein the robot is configured to obey valve constraints which determine in which directions the robot may travel on a path.
17. A robot according to any preceding claim, wherein the robot is configured to perform checks at checkpoint positions on the structure.
18. A system for assembling a structure out of a plurality of structural units, wherein the system comprises at least one robot, preferably a plurality of robots, configured to manoeuvre around the structure.
19. A container for transporting a collection of structural units arranged in layers, wherein at least one layer of the collection of structural units includes fewer structural units than a layer below it, wherein preferably the container comprises an inverse staircase feature which prevents the structural units in a layer with fewer structural units than a layer below it from moving relative to the layer below it.
20. A method of organising at least one robot to assemble a structure out of a plurality of structural units, wherein the method comprises: receiving a model of a structure; determining a structural unit layout model of the structure; and determining a sequence of assembly stages such that the structure is accessible to the at least one robot at all assembly stages.
21. A method of organising at least one robot in an assembly system, wherein the method comprises: setting a valve on a path on which a robot manoeuvres between points in the automatic assembly system, wherein the valve determines the direction of movement the robot may take; communicating the position of the valve to the robot.
22. A method of organising at least one robot in an assembly system, wherein the method comprises: setting a checkpoint on a path on which a robot manoeuvres between points in the automatic assembly system, wherein the robot performs a check when it reaches the checkpoint; communicating the position of the checkpoint to a robot.
23. A staircase system for use with an assembly system comprising at least one robot.
24. A staircase system according to claim 22 or 23, wherein the staircase comprises a plurality of steps between a plurality of access levels, wherein the access levels are at heights corresponding to every three steps.
25. A bypass system for use with an assembly system comprising at least one robot.