DE102017121127A1 - Exploration of an unknown environment by an autonomous mobile robot - Google Patents

Abstract

A method of exploring a robotic mission area by an autonomous mobile robot is described. According to an embodiment, the method comprises starting a reconnaissance trip, wherein the robot detects objects in its vicinity during the reconnaissance trip and stores detected objects as map data in a map while the robot is moving through the robotic use area. During the reconnaissance trip, the robot performs subarea detection based on the stored map data, wherein at least one reference subarea is detected. Subsequently, it is checked whether the reference subarea is completely explored. The robot repeats the partial area detection to update the reference partial area and again checks whether the (updated) reference partial area is fully explored. The exploration of the reference subarea is continued until the test reveals that the reference subarea is fully explored. Subsequently, the robot will continue the reconnaissance trip in a further subarea, provided that a further subarea has been detected, wherein the further subarea is used as the reference subarea.

Description

TECHNICAL AREA

The description relates to the field of autonomous mobile robots, in particular methods for the exploration of a still unknown to the autonomous mobile robot environment in a robotic area.

Hinergrund

Many autonomous mobile robots are available for a wide variety of residential or commercial applications, such as the treatment or cleaning of floor space, the transport of objects or the inspection of an environment. Simple devices can do without the creation and use of a map of the robotic area, for example, by accidentally moving over a floor surface to be cleaned (see eg Publication EP 2287697 A2 from iRobot Corp.). More complex robots use a map of the robotic area, which they create themselves or get provided in electronic form.

Before the map can be used for path planning (and other purposes), the robot must explore its environment in the robotic field to create a map. Methods are known for exploring an environment unknown to the robot. For example, techniques such as Simultaneous Localization and Mapping (SLAM) are used in an exploration phase of a robotic mission. The inventors have set themselves the task to improve the process of exploration of a still unknown to the robot environment within a robotic area.

SUMMARY

The above object can be achieved by a method according to claim 1 or 16 as well as with a robot controller according to claim 19. Different embodiments and further developments are the subject of the dependent claims.

A method of exploring a robotic mission area by an autonomous mobile robot is described. According to an embodiment, the method comprises starting a reconnaissance trip, wherein the robot detects objects in its vicinity during the reconnaissance trip and stores detected objects as map data in a map while the robot is moving through the robotic use area. During the reconnaissance trip, the robot performs subarea detection based on the stored map data, wherein at least one reference subarea is detected. Subsequently, it is checked whether the reference subarea is completely explored. The robot repeats the partial area detection to update the reference partial area and again checks whether the (updated) reference partial area is fully explored. The exploration of the reference subarea is continued until the test reveals that the reference subarea is fully explored. Subsequently, the robot will continue the reconnaissance trip in a further subarea, provided that a further subarea has been detected, wherein the further subarea is used as the reference subarea.

According to a further embodiment, the method comprises starting a reconnaissance trip in a first of a plurality of rooms of the robotic use area, which are connected by doorways. During the reconnaissance trip, the robot detects objects in its environment and stores detected objects as map data in a map while the robot is moving through the robot use area. Furthermore, the robot detects one or more doorways during the reconnaissance trip and checks if the first space is already fully explored. The reconnaissance trip will continue in the first room until the test reveals that the first room is fully explored. Afterwards the exploration trip can be continued in another room.

list of figures

• 1 illustriert einen autonomen mobilen Roboters in einem Robotereinsatzgebiet.
• 2 illustriert anhand eines Blockdiagramms exemplarisch den Aufbau eines autonomen mobilen Roboters.
• 3 zeigt eine von einem mobilen Roboter automatisiert erzeugte Karte seines Robotereinsatzgebietes (einer Wohnung) mit einer Vielzahl von Begrenzungslinien.
• 4 zeigt exemplarisch das Ergebnis eine automatisierten Aufteilung des Robotereinsatzgebietes in Teilgebiete.
• 5 illustriert anhand eines Flussdiagramms ein Beispiel eines Verfahrens zur vollständigen Exploration eines Robotereinsatzgebietes.
• 6 illustriert die Kartierung von Hindernissen durch den Roboter.
• 7 illustriert die Bewegung des Roboters durch ein zu erkundendes Einsatzgebiet.
• 8-13 illustrieren schrittweise die Erkundung eines Robotereinsatzgebietes mit mehreren Zimmern.

Various exemplary embodiments are explained in more detail below with reference to figures. The illustrations are not necessarily to scale and the invention is not limited to the aspects presented. Rather, value is placed on presenting the underlying principles. In the pictures shows:

• 1 illustrates an autonomous mobile robot in a robotic area.
• 2 illustrates exemplarily the structure of an autonomous mobile robot based on a block diagram.
• 3 shows an automated map generated by a mobile robot robot use area (a flat) with a plurality of boundary lines.
• 4 shows by way of example the result of an automated division of the robot application area into subareas.
• 5 Illustrates, by means of a flow chart, an example of a method for the complete exploration of a robotic area.
• 6 illustrates the mapping of obstacles by the robot.
• 7 illustrates the movement of the robot through a field to be explored.
• 8-13 illustrate the step-by-step exploration of a multi-room robotic deployment area.

DETAILED DESCRIPTION

1 exemplifies a cleaning robot 100 as an example of an autonomous mobile robot. Other examples of autonomous mobile robots include service robots, surveillance robots, telepresence robots, etc. Modern autonomous mobile robots navigate card-based, ie they have an electronic map of the robotic operational area. In some situations, however, the robot has no or no up-to-date map of the robotic field and he must explore and map his (unknown) environment. This process is also referred to as "exploration." The robot detects obstacles as it moves through the robotic area. In the example shown, the robot has 100 already parts of the walls W1 and W2 of a room. Methods for exploring the environment of an autonomous mobile robot are known per se. A common method is as mentioned SLAM.

Furthermore, the subdivision of the map created by the robot into subareas is known per se (cf., eg DE 10 2010 017 689 A1 ). The robot subdivides its map according to predetermined criteria, such as door passages detected by sensors, detected floor coverings, etc. The purpose of subdividing the robotic deployment area into several subareas is to create the possibility to individually treat different areas (eg rooms of an apartment). In the case of a cleaning robot, different subregions can be cleaned, for example, with different frequency, with different intensities, at defined times, with certain tools or cleaning agents, etc. However, as long as the robot has not (essentially) completely explored its surroundings, final mapping of the map into subareas is not yet possible There are various strategies for exploring a robotic area, such as random driving, driving along obstacles (especially the outer contour) to complex procedures that determine a next exploration point that the robot can approach and thereby maximize exploration gain (see, eg, D. Lee: The Map-Building and Exploration Strategies of a Simple Sonar-Equipped Mobile Robot (Cambridge University Press, 1996).) However, there is no method that takes into account the particularities of individual room living environments, which are usually connected by doorways and thus clearly separated from one another the device (obstacles) be very complex. As a result, common exploration strategies often move the robot back and forth between rooms, which costs a lot of time and energy. The underlying idea of the invention is to fully explore (explore) a room before the robot changes to the next room. For this purpose, during the exploration, the robot performs a division of the map into subareas in order, for example, to recognize a space as a meaningful subarea. The robot can explore this sub-area and determine when this sub-area (and thus the space) is fully explored.

Before exploring the environment of the robot in more detail, the construction of an autonomous mobile robot will first be briefly described. 2 shows by way of example a block diagram various units (modules) of an autonomous mobile robot 100 , A unit or a module can be an independent module or a part of a software for controlling the robot. A unit can have multiple subunits. The for the behavior of the robot 100 Responsible software may be provided by the control unit 150 of the robot 100 be executed. In the illustrated example, the control unit includes 150 a processor 155 that is trained in a store 156 to execute contained software instructions. Some functions of the control unit 150 can be done at least partially with the help of an external computer. That is, that of the control unit 150 required computing power may be at least partially outsourced to an external computer, which may be accessible via a home network or via the Internet (cloud), for example.

The autonomous mobile robot 100 includes a drive unit 170 , which may include, for example, electric motors, transmissions and wheels, whereby the robot 100 - at least theoretically - can approach every point of a field of application. The drive unit 170 is designed by the control unit 150 received commands or signals in a movement of the robot 100 implement

The autonomous mobile robot 100 further comprises a communication unit 140 to a communication connection 145 to a human machine interface (HMI) 200 and / or other external devices 300 manufacture. The communication connection 145 is for example a direct wireless connection (eg Bluetooth), a local wireless network connection (for example, Wi-Fi or ZigBee) or an Internet connection (for example, to a cloud service). The man-machine interface 200 can give a user information about the autonomous mobile robot 100 For example, in visual or acoustic form output (eg battery status, current work order, map information such as a cleaning card, etc.) and user commands for a work order of the autonomous mobile robot 100 answer. Examples of an HMI 200 are tablet PC, smartphone, smartwatch and other wearables, computers, smart TV, or head-mounted displays, etc. An HMI 200 may additionally or alternatively be integrated directly into the robot, whereby the robot 100 can be operated for example via buttons, gestures and / or voice input and output.

Examples of external devices 300 are computers and servers on which calculations and / or data are outsourced, external sensors that provide additional information, or other household appliances (eg, other autonomous mobile robots) that are used by the autonomous mobile robot 100 work together and / or exchange information.

The autonomous mobile robot 100 can be a work unit 160 Such as a processing unit for processing a bottom surface and in particular for cleaning a bottom surface (eg brush, suction device) or a gripping arm for grasping and transporting objects.

In some cases, such as a telepresence robot or a surveillance robot, another assembly is used to accomplish the intended tasks and is not a unit of work 160 necessary. Thus, a telepresence robot can communicate with the HMI coupled communication unit 140 have, for example, can be equipped with a multimedia unit, which includes, for example, microphone, camera and screen to allow communication between several distant people. A surveillance robot determines unusual events (eg fire, light, unauthorized persons, etc.) on control trips with the aid of its sensors and, for example, informs a control center about this. In this case, instead of the work unit 160 a monitoring unit provided with sensors for monitoring the robotic area.

The autonomous mobile robot 100 includes a sensor unit 120 with various sensors, for example one or more sensors for detecting information about the environment of the robot in its field of application, such as the position and extent of obstacles or other landmarks in the field. Sensors for acquiring information about the environment are, for example, sensors for measuring distances to objects (eg walls or other obstacles, etc.) in the environment of the robot, such as an optical and / or acoustic sensor, by means of triangulation or transit time measurement of an emitted signal Measure distances (triangulation sensor, 3D camera, laser scanner, ultrasonic sensors, etc.). Alternatively or additionally, a camera can be used to collect information about the environment. In particular, when viewing an object from two or more positions, the position and extent of an object can also be determined.

In addition, the robot may have sensors to detect (mostly unintentional) contact (or collision) with an obstacle. This can be realized by accelerometers (which detect, for example, the speed change of the robot in a collision), contact switches, capacitive sensors or other tactile or touch-sensitive sensors. In addition, the robot may have floor sensors to detect an edge in the floor, for example a step. Other common sensors in the field of autonomous mobile robots are sensors for determining the speed and / or the distance traveled by the robot, such as e.g. O-dometer or inertial sensors (acceleration sensor, yaw rate sensor) for determining the position and movement of the robot and Radkontaktschalter to detect contact between the wheel and ground.

The autonomous mobile robot 100 can be a base station 110 be assigned, on which he can load his energy storage (batteries), for example. To this base station 110 can the robot 100 return after completing a task. If the robot has no task left to work on, it can work in the base station 110 wait for a new mission.

The control unit 150 can be designed to provide all the functions that the robot needs to move independently in his field of application and to perform a task. This includes the control unit 150 for example, the processor 155 and the memory module 156 to run a software. The control unit 150 can be based on that of the sensor unit 120 and the communication unit 140 information received control command (eg control signals) for the work unit 160 and the drive unit 170 produce. The drive unit 170 can, as already mentioned implement these control signals or control commands in a movement of the robot. Also in the store 156 contained software can be modular. A navigation module 152 provides, for example, functions for automatically creating a map of the robot deployment area, as well as for planning the movement of the robot 100 ready. The control software module 151 provides eg general (global) control functions and can form an interface between the individual modules.

• • das Erstellen von (elektronischen) Karten durch Sammeln von Informationen über die Umgebung mit Hilfe der Sensoreinheit 120, beispielsweise aber nicht ausschließlich mittels SLAM-Verfahren,
• • das Management von einer oder mehreren Karten zu einem oder mehreren den Karten zugeordneten Einsatzgebieten des Roboters,
• • das Bestimmen der Position und Orientierung (Pose) des Roboters in einer Karte basierend auf den mit den Sensoren der Sensoreinheit 120 ermittelten Informationen der Umgebung,
• • eine kartenbasierte Pfadplanung (Trajektorienplanung) von einer aktuellen Pose des Roboters (Startpunkt) zu einem Zielpunkt,
• • ein Konturfolgemodus, in dem der Roboter (100) sich entlang der Kontur eines oder mehrerer Hindernisse (z.B. einer Wand) in einem im Wesentlichen konstanten Abstand d zur dieser Kontur bewegt,
• • eine Teilgebietserkennung, bei der die Karte analysiert und in Teilgebiete zerlegt wird, wobei beispielsweise Raumgrenzen wie Wände und Türdurchgänge identifiziert werden, wodurch die Teilgebiete die Räume einer Wohnung und/oder sinnvolle Teilbereiche dieser Räume beschreiben.

In order for the robot autonomously to perform a task, the control unit can 150 Functions for the navigation of the robot in his field of use include the above-mentioned navigation module 152 to be provided. These functions are known per se and may include, but are not limited to, one of the following:

• • creating (electronic) maps by gathering information about the environment using the sensor unit 120 but not exclusively by means of SLAM methods,
• The management of one or more cards to one or more application areas of the robot assigned to the cards,
• • determining the position and orientation (pose) of the robot in a map based on that with the sensors of the sensor unit 120 determined information of the environment,
• A map-based path planning (trajectory planning) from a current pose of the robot (starting point) to a destination point,
• A contour following mode in which the robot ( 100 ) moves along the contour of one or more obstacles (eg, a wall) at a substantially constant distance d from that contour,
• • Partial area recognition, in which the map is analyzed and divided into subareas, identifying, for example, spatial boundaries such as walls and doorways, whereby the subareas describe the spaces of a dwelling and / or meaningful parts of these spaces.

The control unit 150 can with the help of the navigation module 152 and based on the information of the sensor unit 120 For example, during operation of the robot continuously update a map of the robotic area, for example, when the environment of the robot changes (obstacle is moved, door is opened, etc.). A current map can then be from the control unit 150 used for short and / or long-term motion planning for the robot. The planning horizon is the path the control unit uses 150 predicts a (target) movement of the robot before it is actually executed. The embodiments described herein relate inter alia to different approaches and strategies for motion planning in certain situations, for example in situations in which certain maneuvers are blocked by obstacles and therefore can not be performed.

In general, one is from the robot 100 usable (electronic) map a collection of map data (eg a database) for storage of location-based information about a field of application of the robot and the environment relevant to the robot in this field. In this context, "location-based" means that the stored information is associated with a position or a pose in a map. A map thus represents a plurality of data records with map data, and the map data may contain any location-related information. Here, the location-related information can be stored in different levels of detail and abstraction, which can be adapted to a specific function. In particular, individual information can be stored redundantly. Often, a collection of multiple maps that relate to the same area but are stored in different form (data structure) is also referred to as "a map".

A technical device is the most useful for a human user in daily life, when on the one hand the behavior of the device is comprehensible and understandable for the user and on the other hand the operation is intuitively possible. It is generally desirable for an autonomous mobile robot (e.g., a floor cleaning robot) to exhibit an intuitively traceable and practical behavior for a human user. For this purpose, the robot must interpret its field of application by technical methods similar and divide into subdivisions as a human user would do (for example, living room, bedroom, hallway, kitchen, dining area, etc.). This allows for easy user-robot communication, for example, in the form of simple commands to the robot (e.g., "clean the bedroom") and / or in the form of notifications to the user (e.g., "cleaning the bedroom completed"). In addition, the mentioned sub-areas can be used for the display of a map of the robotic area and the operation of the robot by means of this card.

Now a division of the robot use area into sub-areas by a user on the one hand according to recognized convention and on the other hand according to personal preferences (and thus user-specific, eg dining area, children's play area, etc.) are made. An example of one well-known convention is the subdivision of an apartment in different rooms such as bedroom, living room and hallway. According to an exemplary user-specific subdivision, a living room could, for example, be divided into a cooking area, a dining area or areas in front of and beside the sofa. The boundaries between these areas can sometimes be defined very "blurred" and are generally subject to the user's interpretation. For example, a cooking area could be characterized by a tile floor, while the dining area is characterized only by the presence of a table with chairs. The adaptation to the human user can be a very difficult task for a robot and often a robot-user interaction may be necessary to correctly divide the robotic area. In order to make this robot-user interaction simple and comprehensible, the map data and the already automatic division must be interpreted and prepared by the device. Furthermore, the human user expects a behavior of the autonomous mobile robot adapted to the division made. Therefore, it may be desired that the subregions may be provided by the user or automatically by the robot with attributes that affect the behavior of the robot.

A technical prerequisite for this is that the autonomous mobile robot has a map of its area of use in order to orient itself in the map. This map is for example built up automatically by the robot and stored permanently. In order to achieve the goal of a user-intuitive division of the robotic area, technical methods are needed which ( 1 ) automatically make a division of the map of the robot use area such as an apartment according to predetermined rules, ( 2 ) permit a simple interaction with the user in order to adapt to the a priori unknown user wishes in the distribution, 3 ) preprocess the automatically generated split to make it easy and understandable to present to the user on a map; and (4) to derive, as far as possible, automatically certain characteristics suitable for achieving user expected behavior.

3 shows a possible representation of a map of a robotic area, as constructed by the robot, for example by means of sensors and a SLAM algorithm. For example, the robot uses a distance sensor to measure the distance to obstacles (eg a wall, a piece of furniture, a door, etc.) and uses the measured data (usually a point cloud) to calculate line segments that define the limits of its area of application. The field of application of the robot can be defined, for example, by a closed chain of line segments (usually a concave, simple polygon), each line segment having a starting point, an end point and consequently also a direction. The direction of the line segment indicates which side of the line segment points to the interior of the mission area, or from which side the robot has "seen" the obstacle indicated by a particular line segment. This in 3 shown polygon describes for the robot the field of application completely, but is very poorly suited for a robot-user communication. A human user may have difficulty recognizing and orienting his own home. An alternative to the mentioned chain of line segments is a raster map, where a grid of, for example, 10 × 10 cm is placed over the robotic area, and each cell (ie 10 × 10 cm box) is marked, if it is occupied by an obstacle. Even such raster maps are difficult to interpret for a human user.

Not only to simplify the interaction with a human user, but also to "work off" the application area in (from the point of view of the user), the robot should first automatically subdivide its robot deployment area into subareas (ie carry out a subarea detection). Such a subdivision into subareas allows the robot to carry out its task more easily, more systematically, more differentiated, (from the point of view of the user) "logical", etc., and to improve the interaction with the user. In order to obtain a meaningful subdivision, the robot must weigh different sensor data against each other. In particular, he can use information about the navigability (difficult / easy) of a section of his field of application to define a sub-area. Furthermore, the robot can assume (refutable) that spaces are usually rectangular. The robot can learn that some changes in the split lead to more meaningful results (so that, for example, certain obstacles are likely to be in a particular sub-area).

As in 3 As shown, a robot is typically capable of detecting obstacles by means of sensors (eg, laser proximity sensors, triangulation sensors, ultrasonic distance sensors, collision sensors, or a combination of the above) and drawing the boundaries of its area of use in the form of boundary lines in a map. However, the limited sensor technology of a robot generally does not allow unambiguous recognition of a subdivision of the application area into different rooms (eg bedroom, living room, corridor, etc.) that is natural for a human user. Even the decision, whether the boundary lines contained in the map (for example, the line between the points J and K in 3 ) belong to a wall or to a piece of furniture is not automatically possible. Even the "border" between two rooms is not easily recognizable to a robot.

In order to solve the mentioned problems and to enable a subdivision of the robotic deployment area into different subareas (eg rooms), the robot sets up "hypotheses" concerning the environment of the robot, which are tested by means of different methods, based on the sensor data. If a hypothesis can be falsified, it is rejected. Are two boundary lines (eg the lines AA 'and OO' in 3 ) approximately parallel and at a distance that corresponds to the usual clear width (there are standardized sizes) of a doorway, the robot can establish the "doorway" hypothesis and conclude that it separates two different rooms. In the simplest case, an automated hypothesis can be tested by the robot "questioning" the user, ie requesting feedback from the user. The user can then either confirm or reject the hypothesis. However, a hypothesis can be tested in an automated way by testing the hypothesis's conclusions for plausibility. If the spaces recognized by the robot (eg by detecting the thresholds) comprise a central space that is smaller than one square meter, for example, the hypothesis that ultimately led to this small central space is probably wrong. Another automated test may be to examine whether or not the two-hypothesis conclusions contradict each other. For example, if six hypotheses regarding a door can be set up and only for five supposed doors can the robot detect a doorstep (a small step), this may be an indication that the hypothesis concerning the door without the door sill is wrong.

When creating a hypothesis by the robot different sensor measurements are combined. For a doorway these are, for example, the passage width, passage depth (given by wall thickness), the existence of a wall to the right and left of the passage or a door projecting into the room. This information can be determined, for example, with a distance sensor by the robot. By means of an acceleration sensor or a position sensor (for example a gyroscopic sensor), it is possible to detect a possibly existing door threshold which the robot passes over. By image processing and a measurement of the ceiling height additional information can be determined.

Another example of a possible hypothesis is the course of walls in the robotic field. These are characterized inter alia by two parallel lines, which are at a distance of a typical wall thickness (see 3 , Thickness dw) and were seen by the robot in two opposite directions (eg the lines KL and L'-K 'in 3 ). However, other objects (obstacles), such as cupboards, shelves, flowerpots, etc., can stand in front of a wall, which can also be identified with the help of hypotheses. A hypothesis can also be based on a different hypothesis. For example, a door is a break in a wall. Thus, if a reliable hypothesis can be made about the course of walls in the field of use of the robot, these can facilitate the recognition of doors and thus the automated subdivision of the robot deployment area.

In order to test and evaluate hypotheses, a degree of plausibility can be assigned to them. In a simple embodiment, a hypothesis is credited with a predefined score for each confirmatory sensor measurement. If in this way a particular hypothesis reaches a minimum score, it is considered plausible. A negative score could lead to the rejection of the hypothesis. In a further developed embodiment, a certain hypothesis is assigned a probability for applying it. This requires a probabilistic model that accounts for correlations between different sensor measurements, but also allows for complex probability statements using stochastic computational models, thus providing a more reliable prediction of user expectations. For example, in certain regions (eg, countries) where the robot is used, door widths may be standardized. If the robot measures such a standardized width, it is most likely a door. Deviations from the standard width reduce the likelihood that it is a door. For example, a probability model based on a normal distribution can be used for this purpose. Another way to create and evaluate hypotheses is to use "machine learning" to build appropriate models and measure functions (see, eg, Trevor Hastie, Robert Tibshirani, Jerome Friedman: "The Elements of Statistical Learning"). 2 , Edition. Springer-Verlag, 2008). For this purpose, for example, map data in different living environments are recorded by one or more robots. These can then be supplemented with floor plans or data entered by a user (eg concerning the course of walls or doorways or a desired partitioning) and evaluated by a learning algorithm.

Another method, which can be used alternatively or in addition to the use of the hypotheses discussed above, is to divide a robotic deployment area (eg, an apartment) into several rectangular areas (eg, rooms). This approach is based on the assumption that rooms are usually rectangular or can be composed of multiple rectangles. In a map created by a robot, this rectangular shape of the room is generally undetectable because in the rooms, numerous obstacles with complex boundaries such as furniture restrict the field of use of the robot.

Based on the assumption of rectangular rooms, the robotic area is covered with rectangles of different sizes intended to render the rooms. In particular, the rectangles are selected such that each point accessible to the robot in the map of the robotic area can be clearly assigned a rectangle. That is, the rectangles usually do not overlap. It is not excluded that a rectangle contains points that are not accessible to the robot (eg because furniture excludes accessibility). Thus, the area described by the rectangles may be larger and of a geometrically simpler shape than the actual robotic area. To determine the orientation and size of the individual rectangles, for example, long straight boundary lines are used in the map of the robotic area, as they occur along walls (see, eg 3 , Straight through the points L 'and K', straight through the points P and P 'and P "and P"'). Various criteria are used to select the boundary lines to use. A criterion may be, for example, that the respective boundary lines to a plurality of other boundary lines are approximately parallel or orthogonal. Another criterion may be that the respective boundary lines are approximately on a straight line and / or comparatively long (ie in the order of the external dimensions of the robotic area). Other criteria for choosing the orientation and size of the rectangles are, for example, recognized doorways or floor coverings. These and other criteria may be used to evaluate them in one or more evaluation functions (analogous to the degree of plausibility of a hypothesis, eg, mapping a score for a hypothesis) to determine the specific shape and location of the rectangles. For example, the boundary lines are given points for satisfied criteria. The boundary line with the highest score is used as the boundary between two rectangles.

Based on the assumption that spaces are essentially rectangular, the robot can choose from the map of the boundary lines (see 3 ) complete the outermost boundary lines to a right-angled polygon (rectilinear polygon). A possible result of a subdivision of a map into subareas is exemplary in 4 shown. Accordingly, the apartment was divided into different rooms (eg bedroom 10 with bed, hall 20 , Living room with open kitchen area 30 ). The subdivision of the apartment into subareas 10 . 20 . 30 (Room) was made eg due to detected doorways. From the subarea 30 (Living room with open kitchen area) was based on recognized flooring continue the subareas 31 (Kitchen area, tiled floor) and 32 (Carpet) (the remaining area 30 has eg a parquet floor).

As already mentioned, a substantially complete map of the robotic area is usually needed before the robot can make a meaningful automatic subdivision of a map into subareas (e.g., rooms). Until then, the robot moves through the robot application area, without being able to pay attention to sub-area boundaries. This can lead to inefficient and difficult to understand behavior of the robot in the "exploration phase". For example, the robot passes through a doorway and enters another room before fully exploring a room, which may result in the robot moving through a large part of the robotic area (eg, a dwelling) and yet at different locations in the dwelling ( in different rooms) "white spots" remain in the map. The robot must finally approach these "white spots" in the map individually to explore them and get a complete map. The exemplary embodiments described here relate inter alia to a method which, at least in some situations, makes the mentioned exploration phase more efficient and makes the behavior of the robot in this exploration phase more "logical" for a user.

5 illustrates an exemplary example of a method for exploring a robotic area by an autonomous mobile robot (cf. 1 , Robot 100 ). A reconnaissance trip can be started at any point in the robotic area ( 5 , Step S10 ), and during the reconnaissance trip, the robot moves through the robotic area. At the beginning of the exploration, the robot has little or no information about its surroundings. The (electronic) map of the robot is recreated and is thus virtually empty at the beginning of the exploration. The robot 100 carries out continuous measurements during the reconnaissance trip ( 5 , Step S11 ) to objects (eg obstacles, also commonly referred to as navigation features) in its environment detect and stores detected objects as map data in a map. For example, walls and other obstacles (eg furniture and other objects) are detected and their position and location are stored in the map. The sensor system used for the detection is known per se and can be used in a sensor unit 120 of the robot 100 be included (cf. 2 ). Calculations necessary for detection can be made by one in the control unit 150 (see. 2 ) of the robot 100 contained processor are performed.

During the reconnaissance trip, the robot leads 100 based on the current map data a partial area detection by (see 5 , Step S12 ), wherein at least one reference sub-area is detected (eg a part of the robotic area in which the robot is currently located). Furthermore, it is checked whether the reference sub-area is completely explored (see 5 , Step S13 ). As long as this is not the case, the robot continues to explore (only) the reference sub-area (see 5 , Step S14 ) by continuously taking measurements to detect objects in the environment of the robot and to store detected objects in the map. The subarea detection is repeated continuously and / or according to predefinable rules in order to update the definition of the reference subarea. The robot then checks again whether the (updated) reference sub-area is completely explored (back to step S13 ). As soon as the test reveals that the (updated) reference subarea is fully explored, it is further checked whether the robotic area is already fully explored (see 5 , Step S15 ). For example, the robot may have detected a sub-area in a sub-area detection, which has not yet been explored. From this, the robot can conclude that the exploration of the robotic area is not yet complete and that the exploration continues in another area ( 5 , Step S16 ), which then becomes the reference sub-area. If the robot can no longer detect another sub-area that has not yet been explored, the exploration of the robotic area is completed.

For updating the reference subarea, both the current map data and the previous limits (that are specified in an earlier subarea detection) of the reference subarea can be taken into account in the subarea detection. The current update of the reference sub-area will be explained in more detail later (see. 8th to 10 ). In one embodiment, for updating the reference sub-area, the already explored part of the robotic area can be subdivided into one or more sub-areas, wherein the reference sub-area is selected according to predefined criteria from the detected sub-areas. If only a single sub-area is detected, this is the reference sub-area. A criterion for selecting the reference subarea may be, for example, the extent of the overlap with the reference subarea previously determined (in the preceding subarea detection). That is, it is selected from the detected sub-areas of that as the current reference sub-area, which has the largest overlap with the previously determined reference sub-area.

Algorithms for the automated division of a map into subareas are known per se. Some of these algorithms only work if the card is fully explored, ie completely enclosed by walls and other obstructions. These algorithms can be used for a not fully explored map when the map is "artificially" completed. For this purpose, for example, a bounding box can be placed around the already explored area, and considered as a "virtual" obstacle. Other ways of completing an incompletely explored map to use the algorithms to automatically map a complete map can also be used as alternatives. In the embodiments described herein, one or more information based on map data may be utilized for the partial area detection, such as the position of walls and / or other obstacles, the position of doorways, and the location of flooring boundaries. Additionally or alternatively, information about the floor structure, the ceiling structure, and / or the wall structure (e.g., stored in the map) may be taken into account. Another criterion for the definition of a subarea are predefinable geometric properties of subareas, such as a minimum or a maximum size of a subarea. Ceiling structures and in particular the corners between the ceiling and a wall can directly provide information about the shape and size of a room. From this, for example, doorways can be recognized as openings in the wall that do not reach to the ceiling. Wall structures, such as windows, can provide clues as to whether the wall is an exterior wall. Floor structures such as, in particular, the change of floor covering or a threshold may be indicia of room boundaries and, in particular, doorways.

Depending on the implementation of the partial area detection, the boundary lines of a partial area can be determined at least partially predictively. In this context, prediction means that existing, already detected and stored in the map contours (eg a wall) are used to predict a boundary line of a subarea. For example, an already detected and stored in the map contour of a wall (virtually) be linearly extended to the Limit of a sub-area to complete. According to another example, a boundary line of a subarea parallel or perpendicular to an already detected and stored in the map contour of a wall (or other obstacle) is set so that it touches the edge of the already explored area of the robotic area. An example of this will be explained later 9 and 10 explained.

The robot tracks its own position in the map. After a partial area detection, the robot can check whether it is still in the (updated) reference subarea. If this is not the case and if the reference sub-area is not yet fully explored, the robot travels back to the reference sub-area to continue the reconnaissance there. The reconnaissance trip will not continue outside the reference subarea. However, if the test indicates that the (updated) reference sub-area has already been completely explored (for example, because it is limited only by contours of obstacles or borderlines to other detected sub-areas), another sub-area will become the reference sub-area and the reconnaissance will continue there.

As mentioned, the sub-area detection may be repeated regularly or in response to the detection of certain events. A repetition of the partial area detection may e.g. are triggered when the robot determines that a certain time interval has elapsed since the last subarea detection, that the robot has traveled since the last subarea detection a certain distance that since the last subarea detection of the explored area of the robotic area has increased by a certain area, or that the effort for the further exploration of the reference sub-area exceeds a certain value. The effort can e.g. be evaluated by means of a cost function. A repetition of the partial area detection may e.g. also be triggered when the robot has reached a destination determined for the exploration. Such a target point can be chosen, for example, on the border between explored and (still) unexplored subareas. While the robot is heading to this point, it can detect new areas with its sensors and thus extend the boundaries of the area explored.

When the robot detects that the reference sub-area has been completely explored, it is saved and its boundaries are not changed later. As long as there are more sub-regions that are not completely explored, the robot selects a further sub-area as reference sub-area and continues the exploration journey there. The previous reference sub-area (or the previous reference sub-areas) can be considered in the further sub-area detection to the effect that its (or their) boundary lines are not changed and thus the boundaries of adjacent sub-areas (at least partially) are fixed. That is, the boundary lines of the previous reference sub-areas can be taken into account in the determination of the boundary lines of other sub-areas.

If the robot is a cleaning robot, it can - after he has completely explored a reference sub-area - clean it before selecting another sub-area as a reference area and continue the reconnaissance trip there. This behavior can be made dependent on a user input, which increases the flexibility of the robot. The robot can therefore accept a user input, depending on which the robot distinguishes three operating modes. The user input (e.g., "explore," "explore and clean," "clean") may be, for example, via an HMI (e.g., on a portable external device or directly on the robot). In a first mode of operation, the robot performs a reconnaissance trip and the robot explores the robotic area and creates a new map. The reconnaissance trip may be implemented according to the methods described herein. In a second operating mode, the robot performs a reconnaissance trip, creating a new map while also cleaning the robotic area. In a third mode of operation, no new map is created, but the robotic area is cleaned based on an existing, stored map. This concept can also be applied to other robots that are not cleaning robots. In this case, the robot performs another activity instead of cleaning.

The robot may mark the already explored area of the robotic area as "explored" on the map (eg, by setting a particular bit or other marker, or by capturing, updating and storing the boundary between explored and unexplored areas). For example, those areas can be marked as explored in the map, which are located at least once within a detection range of a navigation sensor of the robot during the reconnaissance trip (cf. 7 ). The robot can usually detect objects by means of the navigation sensor within the detection range of the navigation sensor, for example by measuring without contact the distance between the navigation sensor and a plurality of points of an object (cf. 6 ).

The robot may terminate the reconnaissance if it determines that the mission has been fully explored (eg, because the explorer is explored) marked area is completely limited by obstacles) and / or if it is not possible to continue the reconnaissance trip because no further (not yet fully explored) subarea has been detected. In this situation, the robot can again perform a partial area detection taking into account the map data concerning the fully explored robotic area. In this final partial area detection, the robot can use other (eg more accurate and / or more elaborate) algorithms than in the repeatedly performed partial area detection during the exploration trip. Alternatively, however, the same algorithm can be used with changed parameters. Finally, after completing the reconnaissance, the robot may return to a starting point where the reconnaissance trip was started or to a base station detected during the reconnaissance trip and stored in the map.

The diagrams in 6 exemplify as an autonomous mobile robot 100 explores his surroundings. In the illustrated example, the sensor unit 120 of the robot 100 (please refer 2 ) a navigation sensor 121 on, which has a defined coverage area Z coverage area. In the example shown, the detection area has Z approximately the shape of a circular sector with the radius d. This situation is shown in diagram (a) of 6 shown. The navigation sensor 121 is designed to contain objects (eg obstacles, furniture and other objects) in the environment of the robot 100 to detect by measuring the distance to the contour of an object, as soon as an object within the detection range Z of the sensor 121 lies. The coverage area Z moves with the robot, and the detection area Z may overlap objects when the robot comes closer to them than the distance d. Diagram ( b ) of the 6 shows a situation in which an obstacle H is within the detection range Z of the navigation sensor 121 of the robot. The robot can be part of the contour of the obstacle H recognize, in the present example this is the line L on the sensor 121 facing side of the obstacle H , The line L is saved in the map. During the exploration, the robot becomes the obstacle H detect from other directions, and so can complete the contour of the obstacle in the map. In the example shown, the detection area has Z the sensors a comparatively small field of view. However, there are also sensors capable of covering a range of 360 °. In this case, the coverage area has Z the shape of a (complete) circle. Other detection areas are possible depending on the sensor and known per se. In particular, the detection area may be a volume, such as an opening cone, especially for sensors for three-dimensional detection of the environment.

The robot 100 usually knows his own position in the map; The robot can change its position 100 for example by means of odometry (eg by means of wheel sensors, visual odometry, etc). Thus, the robot also "knows" which areas of the robotic area it has already explored and can mark these explored areas in the map as "explored". In the in 7 In the example shown, the explored area E comprises all points of the robotic area which, during the reconnaissance trip, at least once in the (with the robot 100 moved) have detected Z detection area. The areas of the map that are not marked as explored can be considered as "white spots" over which the robot has no information (yet). It should also be noted at this point that an object H located in the detection area Z shades off a part of the detection area and effectively makes it smaller (see diagram (b) in FIG 6 ).

The 8th to 10 illustrate the progressive exploration of a multi-room robotic deployment area. 8th shows a situation shortly after the start of a reconnaissance trip. The illustrated robot use area corresponds to that from the 3 and 4 , In the present example, the exploration trip starts in the upper area of the bedroom (cf. 4 , Room 10 ). For the sake of simplicity, it is assumed that the detection range Z of the navigation sensor 121 360 ° and therefore is circular. The circle marked Z indicates the theoretical maximum detection range of the sensor 121 , Accordingly, the boundary EB of the explored area E is given by the detected obstacles around the robot and two circular arcs.

The robot 100 can now perform a partial area detection to a first reference sub-area R (or its limits). The reference subarea R is for example limited on the one hand by the detected obstacles. Second, two virtual tentative boundary lines are defined. Since there are no further clues for their location, they are defined, for example, as straight lines that are orthogonal to the detected obstacles, which are the boundary EB of the explored area e touch (see 8th , horizontal and vertical dashed tangents to the circular arcs EB).

For example, to further explore the reference subarea, the robot may try to one of the boundary lines of the explored area EB to go out, not by one Detected obstacle (eg wall, bed, dresser, etc.) are formed. In this example, the robot moves down into the area of the room below the map 10 while continuing to make measurements and detect and store obstacles in his environment. This situation is in 9 shown. How to get in 9 can see in the cp. to 8th the explored area E increases, with one corner of the wall to the left of the robot the field of view of the sensor 121 of the robot 100 partially shaded. In the situation shown is a part of a doorway (towards the hallway 20 ) in the explored area E. With a renewed partial area detection, other limits result for the reference subarea R. In the subarea detection, further, predefined assumptions can flow in, for example the assumption that a room has a substantially rectangular base area. According to 9 the boundary lines are chosen such that the reference sub-area is a substantially rectangular area that includes the majority of the explored area. The lower boundary line of the reference subarea R (horizontal dashed line in FIG 9 ) is generated, for example, by extending the contour of the wall. The left boundary line of the reference subarea R (vertical dashed line in FIG 9 ) is set, for example, at right angles thereto, so that no boundary line of the explored area EB intersects them (but touches only in the upper area).

Due to the definition of the reference subarea R, the robot moves 100 now not through the doorway in the adjacent room, because that would leave him the reference sub-area. Instead, the robot stays 100 in the reference subarea R and explores the previously unrecognized area (white spot) on the lower left (in the map) and the explored area E continues to grow and now covers almost the entire room (see 2 , Bedroom 10 ). This situation is in 10 shown. During the renewed partial area detection, the robot recognizes 100 the doorway to the adjacent room (see 4 , Hall 20 ), as a line D is entered in the map and is used to define a boundary between two subregions. Therefore, the robot becomes the already explored area on the other side of the line D another (new, not yet detected) subarea T assign. In the in 10 the partial area detection thus provides two partial areas, the reference partial area R and the further subarea T ,

Starting from the in 10 As shown, the robot does not move directly to the next sub-area T drive (although it is very close to the robot 100 lies), to continue the exploration, but first in the reference area R Explore the existing "white spot" (in the map on the top left). After that, the robot will notice that the reference subarea R is completely explored, as it is exclusively made of obstacles (eg walls, furniture, etc.) or a boundary line between two subregions (the line D , Doorway) is limited. Only then does the robot move into the further subarea T , which is defined as a new reference sub-area, and continues there his reconnaissance trip.

11 shows the robot 100 in the exploration of the sub-area T, which is now the reference sub-area. After the robot 100 has passed through the doorway, he is left bent in the example shown and the corridor (see. 4 , Hall 20 ) in the direction of the living room (cf. 4 , Living room 30 ) followed. In 11 are again the limits EB of the explored area e The boundary lines on the right and left sides of the reference subarea T are straight lines, which run perpendicular to a detected wall and affect the boundaries EB of the explored area. In the illustration according to 11 the explored area E already extends into the adjacent room (cf. 4 , Living room 30 ), where the robot 100 the doorway to the living room has not yet recognized as such. That's why the robot is set 100 the reconnaissance trip continues into the neighboring room and comes into a situation that in 11 is shown.

In the in 11 the example shown recognizes the robot 100 the door passage D 'only after he has driven through this. A subsequent partial area detection leads (due to the now recognized doorway D ') to the detection of a further subarea U , In this situation, the robot can detect that it is no longer in the reference sub-area T Although this is not yet fully explored (left is the reference sub-area is not yet limited by a detected obstacle or another already recognized sub-area and therefore explored incomplete). The robot 100 will therefore return to the reference sub-area T and explore this further.

In the in 13 the situation presented has the robot 100 also explores the left-hand area of the reference sub-area T, and the robot will detect that the reference sub-area T now fully explored, as it is exclusively detected contours of obstacles (walls) and detected adjacent sub-areas R and U is limited. Starting from the in 13 the situation is now the robot becomes the sub-area U choose as a new reference sub-area (because this is not yet completely explored) and continue the reconnaissance trip there.

In a simple variant of the method described here, the partial area detection is reduced to a detection of the doorways. The detection of an (open) doorway implies the detection of an adjacent room. That is, the sub-area detection detects during the reconnaissance trip practically only different rooms than different sub-areas. Incidentally, this embodiment is the same as or similar to the above-described embodiments. The robot will first fully explore a room before going through a doorway to continue exploration in a neighboring room. If the robot accidentally moves into a neighboring room, for example because a doorway is only recognized as such after the robot has detected this doorway as such, the robot can determine that it has left the previously explored space, although it has not yet fully explored is. In this situation, the robot will end the reconnaissance trip in the room in which it is currently located and drive through the previously accidentally passed doorway again (in the opposite direction), in this way return to the previously explored space and explore this further.

The methods described herein may be implemented in the form of software. The software may be executed on a robot, on a human-machine interface (HMI) and / or on any other computer such as, in particular, a home server or a cloud server. In particular, individual parts of the method may be implemented by means of software which may be subdivided into different software modules and which may be executed on different devices. When the robot "does something" (eg, performs a process step), this process (eg, process step) may be performed by the control unit 150 been initiated (cf. 2 ). This control unit 150 can (possibly together with other units) form an autonomous mobile robot control unit that controls all functions and behaviors of the autonomous and mobile robot (using, among other things, other units and modules present in the robot).

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• EP 2287697 A2
• DE 102010017689 A1 [0009]

Claims (19)

A method of exploring a robotic area by an autonomous mobile robot, the method comprising: Starting a reconnaissance trip, during which the robot performs the following during the reconnaissance trip: Detecting objects in the environment of the robot and storing detected objects as map data in a map while the robot is moving through the robotic area; Performing a partial area detection based on the stored map data, wherein at least one reference partial area is detected; Check that the reference subarea is fully explored; Repeating the partial area detection to update the reference partial area and repeating the check as to whether the reference partial area is fully explored, and continuing the exploration of the reference partial area until the examination reveals that the reference partial area is fully explored; and subsequently Continue the reconnaissance trip in a further subarea, provided that a further subarea has been detected, the further subarea being used as reference subarea. The method according to Claim 1 , wherein for updating the reference sub-area, the sub-area detection takes into account both the current map data and the previous boundaries of the reference sub-area. The method according to Claim 2 in which, for updating the reference sub-area, the part of the robot-use area which has already been explored is subdivided into one or more sub-areas, the reference sub-area being selected from the detected sub-areas according to predefined criteria. The method according to Claim 3 , Which is selected from the detected sub-areas that as the reference sub-area, which has the largest overlap with the previously determined reference sub-area. The method according to one of Claims 1 to 4 , wherein for the sub-area detection at least one of the following based on the map data information is used: position of walls and / or other obstacles, position of doorways, floor coverings, floor structure, ceiling structure, wall structure, specifiable geometric properties of a sub-area, such as a minimum or a maximum size of a sub-area. The method according to one of Claims 1 to 5 , wherein the boundary lines of a sub-area are determined at least partially predictive in the sub-area detection, in particular by contours of already recognized objects are extended and used as boundary lines and / or by using predefined criteria determined lines are used as boundary lines that already explored that part of the robotic field was, touch. The method according to one of Claims 1 to 6 , wherein the robot checks after a partial area detection, if he is still in the reference sub-area, and, if not, he goes back into this. The method according to one of Claims 1 to 7 wherein the partial area detection is repeated when the robot detects at least one of the following events: a certain time interval has elapsed since the last partial area detection, the robot has traveled a certain distance since the last partial area detection; since the last subfield detection, the explored area of the robotic field has increased by a certain area; the cost of further exploration of the reference sub-area exceeds a certain value, the effort is evaluated with a cost function; and the robot has reached a target point for exploration. The method according to one of Claims 1 to 8th wherein a reference sub-region is stored and no longer altered after the robot has determined that it is fully explored. The method according to Claim 9 in which - if the exploration is continued in a further subarea as a reference subarea - the previous reference subarea (s) are taken into account in the further subarea detection in such a way that its (their) boundary lines no longer change and / or that its (their) whose) boundary lines are considered for the determination of boundary lines of further subregions. The method according to one of Claims 1 to 10 wherein the robot is a cleaning robot and cleans a reference sub-area after determining that it is fully explored and before continuing the exploration in a further sub-area as a reference sub-area. The method according to one of Claims 1 to 11 wherein the robot marks those areas of the robot sub-area as explored in the map, which are located at least once within a detection range of a navigation sensor of the robot during the reconnaissance trip. The method according to Claim 12 wherein the robot detects objects by means of the navigation sensor within the detection range of the navigation sensor by non-contact measurement of the distance between the navigation sensor and a plurality of points of an object. The method according to one of Claims 1 to 13 which further includes terminating the exploration trip if it is not possible to continue the reconnaissance trip because no further subarea was detected; and re-performing a partial area detection based on the map data regarding the fully explored robotic area. The method according to one of Claims 1 to 13 which further includes terminating the exploration trip if it is not possible to continue the reconnaissance trip because no further subarea was detected; and returning to a starting point where the robot is started or starting a base station that was detected during the reconnaissance trip. A method of exploring a robotic area by an autonomous mobile robot, the method comprising: Commencing a reconnaissance trip in a first of several rooms of the robotic area connected by doorways, the robot performing the following during the reconnaissance trip: Detecting objects in the environment of the robot and storing detected objects as map data in a map while the robot is moving through the robot use area; Detecting one or more doorways; Checking that the first room is already fully explored and continuing the exploration trip in the first room until the examination reveals that the first room is fully explored; and then continue the reconnaissance trip in another room. Method according to Claim 16 wherein a doorway may be detected before, during, or after it has been passed, and wherein if the robot has traversed a doorway before the first room has been fully explored, the doorway passes in an opposite direction to return to the first room , A method for controlling an autonomous mobile robot, comprising: accepting a user input by the autonomous mobile robot, determining an operating mode according to the user input, wherein the robot is in a robotic area and in a first operating mode the robotic area according to a method according to any one of Claims 1 to 17 explores and recreates a map, in a second mode of operation, the robotic area according to a method according to one of Claims 1 to 17 exploring thereby recreating a map and performing an action, in particular cleaning the robotic area, and performing a task in the robotic area in a third operating mode, in particular cleaning the robotic area, using a previously created and stored map for navigation in the robotic area. A robot controller for an autonomous mobile robot, comprising: a memory for storing software instructions; at least one processor configured to execute the software instructions, wherein when the software instructions are executed, the processor causes the robot controller to perform a method according to any one of Claims 1 to 17 perform.

Images