JP6037608B2 - Service control system, service system - Google Patents

Description

  The present invention relates to a system that provides a service in which a machine grasps the state of the real world and autonomously performs physical operation or information processing.

  A robot is a typical system in which a machine grasps and recognizes the state of the real world and the machine makes a useful action to the real world based on the recognition result. In Patent Document 1 below, in order to move an autonomous mobile robot according to a desired route without requiring installation of rails or the like, its own position is recognized from the amount of movement of the robot, and the route is determined based on map information according to the travel rules. A method for searching and controlling the robot along the path is shown.

  On the other hand, a system that interacts with a person can be positioned as one of systems in which a machine works based on the state by regarding the person as a component of the real world. In Patent Document 2 below, a method is described in which an action from a user is input and an action determined based on the input is performed on the user. In this document, the behavior of the user interface is gradually increased so that a reaction suitable for the user's action is performed by updating the data structure held in the machine by a predetermined method according to the input action. It has a mechanism to evolve and update.

JP 2010-191502 A JP 2007-265433 A

  In Patent Document 1, since the rules for searching for a route are constant, an appropriate moving route is not always determined or determined when the environmental condition in which the robot moves or when the moving purpose is changed. The movement along the route is not always performed accurately. In other words, it is desirable that the method of determining the operation and the control method change so as to be appropriate with changes in various external conditions such as the object and purpose.

  The above-mentioned patent document 2 shows an updating means for an action determination method in which an action that becomes a reaction becomes more appropriate depending on how an input operation is performed by a user, and occurs with a change in the external condition described above. This is one solution to the problem. However, in the present technology, since the means for accepting the user input operation for the user interface and the means for accepting the user input operation as information for updating the action determination method are the same, the information obtained thereby is It is not always sufficient as information for updating the action determination method. In other words, the information necessary for the system to respond more appropriately, in other words, the information necessary for the system to evolve and update, is limited to the information that can be obtained by the system itself. That is, it is not evaluated from the viewpoint of the third party. For this reason, it is difficult to accurately determine whether the above evolution / update is actually appropriate.

  The present invention has been made in view of the above problems, and in a service system in which a machine makes a useful action on the real world based on the real world state, even when environmental conditions, purposes, etc. change, The purpose is to change the approach to be more appropriate.

  The service control system according to the present invention includes a second observation unit that observes a physical environment different from the first observation unit included in the real world interface system, and the real world interface system operates based on the observation result of the first observation unit. The operation of the real world interface system is adjusted so that the difference between the case and the case of operating based on the observation result of the second observation unit becomes small.

  According to the service control system of the present invention, in a service system in which a machine or the like makes a useful action on the real world based on the state of the real world, even if environmental conditions or service purposes change, the second observation unit Based on objective measures from observations, the operation of the real-world interface system can be adapted so that the approach is more appropriate from a system requirements perspective.

It is a functional block diagram explaining the basic concept of the service system 1000 which concerns on this invention. 2 is a functional block diagram of a service system 1000 according to Embodiment 1. FIG. It is a figure which shows the specific example of the distance sensor. It is a figure which shows the example of the sensing data measured using the distance sensor. It is a figure which shows the example of the map data. It is a figure which shows the outline | summary of a matching process. 6 is a diagram illustrating an example of route data 233. FIG. 5 is a flowchart illustrating a procedure for updating map data 231 by the service system 1000 according to the first embodiment. 6 is a functional block diagram of a service system 1000 according to Embodiment 2. FIG. 10 is a flowchart illustrating a procedure for updating route data 233 by the service system 1000 according to the second embodiment. FIG. 10 is a functional block diagram of a service system 1000 according to a third embodiment. It is a flowchart explaining the procedure in which the service system 1000 which concerns on Embodiment 3 updates the algorithm and parameter for producing a conveyance plan.

<Basic concept of the present invention>
FIG. 1 is a functional block diagram illustrating the basic concept of a service system 1000 according to the present invention. The service system 1000 includes a service control system 100 and a real world interface system 200. The real world interface system 200 is a system that provides services by performing physical operations or information processing. The service control system 100 is a system that controls the operation of the real world interface system 200.

  The real world interface system 200 includes a first observation unit 210, an operation unit 220, and an operation instruction unit 230. The first observation unit 210 observes the physical environment around the real world interface system 200 and outputs the observation result to the operation instruction unit 230 as first observation information. The operation unit 220 includes functional units that perform operations such as a robot that provides a service based on a physical operation such as package transportation and a computer that provides an information processing service. The operation instruction unit 230 outputs operation instruction information based on the observation result of the first observation unit 210 and controls the operation of the operation unit 220.

  The service control system 100 includes a second observation unit 110 and an update instruction unit 120. The second observation unit 110 observes a physical environment different from that observed by the first observation unit 210 around the real world interface system 200 and outputs the observation result to the update support unit 120 as second observation information. The update instruction unit 120 updates control parameters, algorithms, and the like when the operation instruction unit 230 controls the operation unit 220 using the information used internally by the operation instruction unit 230 and the second observation information.

  The first observation unit 210 is a functional unit for adjusting the service operation by observing the surrounding situation by itself when the real world interface system 200 provides the service. On the other hand, the second observation unit 110 is a functional unit that observes the operation of the real world interface system 200 from a different viewpoint from the first observation unit 210 from the outside of the real world interface 200, and is objective for the real world interface 200. It can be said that this is a functional unit that performs operation monitoring by a third party.

  In this way, in addition to the subjective information acquired by the system providing the service itself, information that cannot be obtained by subjective observation alone is obtained by using objective observation results by an external third party in combination. Therefore, it is considered that the service operation of the operation unit 220 can be updated more appropriately.

  For example, when the observation result by the first observation unit 210 and the observation result by the second observation unit 110 are different, the algorithm of the operation instruction unit 230 is updated using either one of the observation results, and there is a gap between the two observation results. It is conceivable to prevent this from occurring. Alternatively, when the operation result is different between the case where the operation unit 220 operates based on the first observation information and the case where the operation unit 220 operates based on the second observation information, the control is performed so that no difference occurs between the two operation results. Applications that correct the parameters are conceivable.

  It should be noted that which of the observation results from the first observation unit 210 and the observation results from the second observation unit 110 should be given priority depends on the nature of the service provided by the service system 1000. For example, in the case of a service that gives priority to overall optimization, the second observation unit 110 observes the entire service system 1000 from a bird's-eye view, and the first observation unit 210 fine-tunes the operation based on observation results at a relatively short distance. Possible uses.

  The concept of the service system 1000 according to the present invention has been described above. Hereinafter, a specific embodiment of the service system 1000 according to the present invention will be described.

<Embodiment 1>
FIG. 2 is a functional block diagram of the service system 1000 according to the first embodiment of the present invention. The service system 1000 according to the first embodiment is a system that provides a service for transporting goods in a distribution warehouse or the like using an autonomous transport vehicle that can automatically reach a destination without being operated by a person. .

  In the first embodiment, the real-world interface system 200 is a system in which an autonomous transport vehicle transports luggage. The first observation unit 210 is configured as a distance sensor, and the operation unit 220 is configured as an autonomous transport vehicle. The operation instruction unit 230 includes map data 231, a position estimation unit 232, route data 233, and a movement control unit 234.

  The distance sensor 210 measures the distance between the autonomous transport vehicle 220 and its surrounding objects, and outputs the distance as sensor data 211 to the position estimation unit 232. The autonomous transport vehicle 220 transports luggage and the like according to instructions from the movement control unit 234.

  The map data 231 is data describing an actual measurement map of a place where the autonomous transport vehicle 220 transports luggage and the like. The position estimation unit 232 estimates the position and orientation of the autonomous guided vehicle 220 by comparing and comparing the distance sensor data 211 and the map data 231, and outputs the estimated position and orientation to the movement control unit 234. The route data 233 is data describing a locus that the autonomous guided vehicle 220 should move on the map data 231. The movement control unit 234 controls the operation of the autonomous transport vehicle 220 based on the estimation result of the position estimation unit 232 such that the autonomous transport vehicle 220 moves according to the route data 233. Control of the autonomous guided vehicle 220 will be described later.

  In the first embodiment, the service control system 200 is a system that updates and controls the movement route of the autonomous guided vehicle 220 by updating the map data 231. The second observation unit 110 is configured as a wide area distance sensor. The update instruction unit 120 includes a terrain detection unit 121 and a map update unit 122.

  The wide-area distance sensor 110 is a distance sensor having a wider measurement range than the distance sensor 210, and measures the terrain where the autonomous guided vehicle 220 moves over a wide area, and outputs the result as wide-area distance sensor data 111. The terrain detection unit 121 detects the terrain where the autonomous guided vehicle 220 moves based on the wide-area distance sensor data 111. The map update unit 122 updates the map data 231 to the latest state using the detection result of the landform detection unit 121 and the distance sensor data 211. This update process will be described later.

  The configuration of the service system 1000 according to the first embodiment has been described above. Hereinafter, a method for moving the autonomous guided vehicle 220 according to the route data 233 using the distance sensor 210 and the map data 231 will be described.

  FIG. 3 is a diagram illustrating a specific example of the distance sensor 210. The distance sensor 210 measures the distance between the peripheral object 300 and the distance sensor 210 in a plurality of directions around the distance sensor 210. Various specs of the distance sensor 210 can be considered. Here, as shown in FIG. 3, on a horizontal plane at a certain height, in the direction of −90 ° to + 90 ° with respect to the front of the distance sensor 210. It is assumed that distance data to the peripheral object 300 can be measured instantaneously for a total of 181 different directions for every 1 °. It is assumed that the distance sensor 210 is mounted on the autonomous transport vehicle 220.

  FIG. 4 is a diagram illustrating an example of sensing data measured using the distance sensor 210. Each sensing data is expressed as (φ0, d0), (φ1, d1),..., (Φi, di),. The subscript i for each variable indicates a data number, φi indicates the i-th measurement direction, and di indicates the distance to the peripheral object 300 in the φi direction. Since the distance sensor 210 is assumed to be capable of measuring in a total of 181 different directions every 1 ° in the direction of −90 ° to + 90 ° with respect to the front, φ0 = −90 °, φ1 = −89 °,. .., φ180 = + 90 °.

  FIG. 5 is a diagram illustrating an example of the map data 231. The map data 231 represents the presence of the peripheral object 300 on the cross section measured by the distance sensor 210 as a digital image as shown in FIG. In this image, the boundary portion of the peripheral object is expressed in black.

  It is assumed that the autonomous guided vehicle 220 moves on a flat ground. The degree of freedom related to the position and orientation that changes due to the movement can be expressed by two parameters (x, y) for the position, one parameter for θ for the direction angle of the autonomous guided vehicle 220 facing the orientation, and a total of three parameters. The coordinate system representing these parameters is assumed to coincide with the coordinate system representing the map data 231 as shown in FIG.

  Estimating the position and orientation of the autonomous guided vehicle 220 means obtaining the three parameters (x, y, θ) from the sensing data of the autonomous guided vehicle 220 in an arbitrary position and orientation. Several methods can be considered as means for realizing this function. Here, a method based on matching processing, which is the most basic method, will be described.

  The method based on the matching process is to superimpose the measured sensing data on a map at various positions and orientations (directions), find the most applicable position and orientation, and use this to estimate the position and orientation parameters of the autonomous guided vehicle 220 It is said that. The principle is based on the idea that when the autonomous transport vehicle is in its position and orientation, the same data as the measured data will be obtained with the highest probability.

  FIG. 6 is a diagram showing an outline of the matching process. When the map data 231 and the distance sensor data 211 are given, the position estimation unit 232 obtains the position and orientation of the autonomous guided vehicle 220 that is determined to be most applicable by the above-described method. The translation component (x, y) and the rotation component θ at this time are the position and orientation estimation values of the autonomous guided vehicle 220.

  Next, route data 233 that defines the travel route that the autonomous guided vehicle 220 should proceed will be described. The route data 233 is obtained by numerically converting the trajectory that the autonomous guided vehicle 220 should pass on the map as a straight line or a curved line.

  FIG. 7 is a diagram illustrating an example of the route data 233. The start point 701 indicates the position where the autonomous guided vehicle 220 is currently located, and the goal point 702 indicates the position that is the destination. A curve 703 connecting the start point 701 and the goal point 702 becomes a movement route of the autonomous guided vehicle 220. As a specific description method for describing a movement route as route data 233, points are plotted at high density on a straight line or a curve representing the route, and the coordinate values of the points on the map are (x0, y0), ( It is conceivable to sequentially express in the format of x1, y1), ..., (xe, ye). (X0, y0) indicates the start point 701, and (xe, ye) indicates the goal point 702.

  The position estimation unit 232 sequentially obtains the position and orientation of the autonomous guided vehicle 220 by the above-described position and orientation estimation method. The movement control unit 234 determines the moving direction and moving speed of the autonomous guided vehicle 220 by comparing the position and orientation estimation results with the route data 233, and controls the moving device of the autonomous guided vehicle 220 based on this. The autonomous guided vehicle 220 is moved.

  The service control system 100 has a function of updating the map data 231. In general distribution warehouses, not only the movement of goods, but also the rearrangement of shelves, various distribution equipment, etc. is carried out at a certain frequency. Therefore, even the map data 231 once created may not accurately represent the actual shape with time. In order to cope with this, the service control system 100 modifies the map data 231 to be the actual one. The specific method will be described below.

  The wide-range distance sensor 110 can correctly recognize the current three-dimensional shape of the place where the autonomous guided vehicle 220 moves, even if the terrain is several hundred meters away. The topography can be measured with an error of about 1 cm at a spatial sampling rate of about. Since such a wide-range distance sensor is already generally used at a commercial level, a detailed description of the mechanism and the like is omitted.

  The wide-area distance sensor 110 measures the three-dimensional shape of the environment in which the autonomous transport vehicle 220 such as a distribution warehouse moves from one place or several places by a human operation or a preset automatic operation. The three-dimensional shape of the environment measured by the wide-range distance sensor 110 is cut by a certain cross section, that is, the height plane to which the distance sensor of the autonomous guided vehicle 220 is attached, and is almost equivalent to the map data 231. .

  However, the shape information of the environment acquired by this method has a defect that there are many missing parts compared to the map data 231. For this reason, the above-described cross-section information cannot be directly used as the map data 231. This is because the sampling rate of each point to be measured is a few centimeters as described above, but it is sparse to use as a map, and an object in front of the wide-range sensor 110 in the environment. This is due to problems such as being unable to measure an object or a wall behind it.

  Therefore, in the first embodiment, the terrain detection unit 121 creates terrain data related to the map data 231 based on the cross-sectional information acquired by the wide-area distance sensor 110. The terrain data is a condition for creating the map data 231 so that the map data accurately represents the actual situation, such as the presence or absence of an object in a certain place. Specifically, each place is classified into three states: (1) an object exists, (2) an object does not exist, and (3) an unknown. It is done.

  The map update unit 122 updates the map data 231 by the following procedure using the above-described terrain data, the distance sensor data 211 obtained while the autonomous guided vehicle 220 is moving, and the current map data 231.

  The map updating unit 122 first compares the terrain data with the distance sensor data 211, and obtains the position and orientation of the autonomous guided vehicle 220 that the distance sensor data 211 applies to the terrain data without contradiction. This is based on the same idea as the method for obtaining the position and orientation of the autonomous guided vehicle 220 by matching the map data 231 and the distance sensor data 211 described above. Specifically, the distance sensor data 211 is placed at various positions and orientations, and is compared with the terrain data so that the object is measured where there is no original object, or there is no object where the original object exists. It is conceivable to set the position and orientation to be obtained with the position and orientation with the least contradiction, for example, measured at the same time. More specifically, a method is conceivable in which an evaluation value to which an appropriate value is added every time a contradiction occurs is defined and a position / posture at which the lowest evaluation value is obtained is determined.

  Next, the map updating unit 122 compares the distance sensor data 211 matched with the obtained position and orientation with the current map data 231 and corrects the map data 231. Specifically, if a point on the map data 231 that is determined to have an object on the distance sensor data 211 is a blank point (a point where no object exists), that point is used as an existence point (an object exists). If a point on the map data 231 that is determined to have no object on the distance sensor data 211 is an existing point, a method of correcting it to a blank point can be considered.

  The update instruction unit 120 sequentially performs the above processing on the distance sensor data 211 obtained while the autonomous guided vehicle 220 moves. Thereby, the map data 231 is updated throughout. Since the distance sensor 210 can continuously obtain the above-mentioned 361 data every time the autonomous guided vehicle 220 moves, the number of points that can be measured is extremely large and dense when the autonomous guided vehicle 220 continues to move. become. Furthermore, even if it is an object that cannot be measured because it is hidden behind a shield from a certain location, it can be measured if the autonomous guided vehicle 220 moves near the shield.

  FIG. 8 is a flowchart illustrating a procedure for updating the map data 231 by the service system 1000 according to the first embodiment. This flow can be performed, for example, when the real world interface system 200 does not provide a service. Hereinafter, each step of FIG. 8 will be described.

(FIG. 8: Steps S801 to S802)
The wide area distance sensor 110 measures the terrain over the entire environment where the autonomous guided vehicle 220 moves (S801). The update instruction unit 120 creates cross-sectional information that is a cross-sectional view of the environment based on the wide-area sensor data 111 (S802).

(FIG. 8: Steps S803 to S804)
The landform detection unit 121 creates the above-described landform data based on the cross-sectional information obtained in step S802 (S803). The update instruction unit 120 instructs the movement control unit 234 to move the autonomous transport vehicle 220 along the route data 233, and sequentially acquires the distance sensor data 211 (S804).

(FIG. 8: Step S805)
The map updating unit 122 compares the terrain data and the distance sensor data 211, and updates the current map data 231 according to the above-described procedure.

<Embodiment 1: Supplement>
As a data communication method in the first embodiment, the map update unit 122 is mounted on the autonomous transport vehicle 220, and data is transmitted and received through the wireless network between the wide-area distance sensor 110 and the map update unit 122, the map update unit 122. May be placed outside the autonomous guided vehicle 220, and data may be transmitted and received between the map update unit 122 and the autonomous guided vehicle 220 through a wireless network.

  For example, the flow described in FIG. 8 is performed in actual operation by running the autonomous guided vehicle 220 in the sky (empty) every time a certain period passes, or measurement by the wide-range distance sensor 110 is performed. It is conceivable that the distance sensor data 211 is temporarily saved as log data while the autonomous guided vehicle 220 moves back and forth, and the map data is updated using these data later.

<Embodiment 1: Summary>
As described above, the service system 1000 according to the first embodiment uses the wide-area distance sensor 110 and the distance sensor 210 together, and updates the map data 231 by comparing and comparing the results measured by them. Thereby, since the map data 231 can be corrected using observation results based on a plurality of viewpoints, the real world interface system is compared with the case where a service is provided based on the observation results of the real world interface system 200 itself. The operation of the autonomous guided vehicle 220 suitable for the service quality required for the service provided by the 200 can be more appropriately maintained.

  In addition, the service system 1000 according to the first embodiment complements the problem that the observation point becomes sparse and the problem that the defect point is generated due to the shielding object that occurs when the wide-range distance sensor 110 is used alone by the distance sensor 210. can do.

<Embodiment 2>
In the second embodiment of the present invention, as in the first embodiment, an autonomous guided vehicle used in a distribution warehouse or the like is targeted. Here, a procedure for setting the route data 233 will be described.

  In the second embodiment, the person or the system sets the transport route as route data 233 in advance. A plurality of transfer routes are set, and an operator or system selects an appropriate route when carrying out the transfer so that the autonomous guided vehicle 220 moves along the desired route. The control method for moving the autonomous guided vehicle 220 according to the route data 233 is as described in the first embodiment.

  When the map data 231 is updated or the environment shape changes due to movement of an object placed in the surrounding environment, the autonomous guided vehicle 220 moves along a different track from the route data 233 set in advance. This is because, for example, the relative positional relationship between the route data 233 and the map data 231 set in the past is shifted. Therefore, in the second embodiment, the movement locus of the autonomous guided vehicle 220 is observed from the outside, the difference between the actual movement locus and the set route data is obtained, and the route data 233 is corrected. As a correction method, the route data 233 is basically moved in a direction to cancel the generated deviation.

  FIG. 9 is a functional block diagram of the service system 1000 according to the second embodiment. The description of the configuration similar to that of the first embodiment will be omitted as appropriate, and the description will focus on the parts different from the first embodiment.

  In the second embodiment, the update instruction unit 120 includes a comparison unit 123 and a route update unit 124. These operations will be described later.

  In the second embodiment, the second observation unit 110 is configured as a trajectory observation unit that observes the movement of the autonomous guided vehicle 220 from the outside. Specifically, a ceiling camera installed on the ceiling of a mobile environment can be considered as an example. By installing a large number of trajectory observation units 110 where the autonomous guided vehicle 220 moves, it is possible to capture all areas where the autonomous guided vehicle 220 can move. The trajectory observation unit 110 recognizes the position of the autonomous guided vehicle 220 based on the observation result.

  The technique for recognizing the position of the autonomous guided vehicle 220 is realized by means for extracting a part of the autonomous guided vehicle 220 and means for obtaining the position. The means for extracting a part of the autonomous guided vehicle 220 is a method of pasting a mark of a predetermined shape or color on the autonomous guided vehicle 220 and finding it by image processing by template matching, or the shape or color of the autonomous guided vehicle 220. Is stored in advance as a transport vehicle model, and the upper surface portion of the autonomous transport vehicle 220, that is, the portion visible from the ceiling camera, is extracted by comparing the model with a captured image. As a method of recognizing the position of the autonomous guided vehicle 220, it is conceivable to calculate coordinate values in the environment for a specific position such as the extracted marker or the center of gravity of the upper surface portion of the autonomous guided vehicle 220. In the following description, this point will be referred to as a target point. As a method for calculating the coordinate value of the target point, since the autonomous guided vehicle 220 moves in a plane, the height of the target point from the ground is constant, and an intersection of the plane representing the height and the direction from the camera toward the target point is determined. There are a method for obtaining the three-dimensional position of the target point and a method for obtaining the three-dimensional position of the extracted target point using all the above-mentioned ceiling cameras as stereo cameras. In any method, by fixing camera internal parameters such as the camera installation position and installation direction and zoom, and performing camera calibration to obtain those values in advance, the above three-dimensional coordinate values can be obtained from the environment. In the same coordinate system. The position of the target point thus obtained is set as the position of the autonomous transport vehicle 220.

  The trajectory observation unit 110 can obtain the trajectory data 112 that represents the actual trajectory of the autonomous guided vehicle 220 as a numerical value by obtaining the position of the autonomous guided vehicle 220 as a time series by the method described above. Next, a method for correcting the route data 233 using the locus data 112 will be described.

  As described above, the route data 233 is set as a curve on the map. Consider that the trajectory data 112 obtained from the observation values of the trajectory observation unit 110 is drawn on the route data 233 in an overlapping manner. At this time, there is a possibility that the route data 233 and the trajectory data 112 do not overlap with each other due to the above-described accumulation of errors. In this case, the autonomous guided vehicle 220 is moving along a different path from the set route. It is considered that the difference between the locus indicated by the route data 233 and the actually observed locus indicates an approximate difference between the intended route and the actual route.

  The comparison unit 123 calculates a difference between the route data 233 and the trajectory data 112. The route update unit 124 corrects the route data 233 in the direction opposite to the above difference. As a result, even if there is a difference between the preset route data 233 and the actual trajectory data 112, the autonomous guided vehicle 220 moves along the originally intended route. However, since the generated difference is not necessarily the same as the correction amount, the correction amount is not limited to the same method as the absolute value of the difference, and the correction amount is obtained by multiplying the absolute value of the difference by a value between 0 and 1. It is conceivable that the correction amount is converged so as to be close to the intended path by repeating observation and correction. In addition, when the absolute value of the difference is larger than a predetermined threshold value, it is conceivable to introduce a practical countermeasure such as not performing correction but notifying the system administrator as an abnormal state.

  FIG. 10 is a flowchart illustrating a procedure for updating the route data 233 by the service system 1000 according to the second embodiment. This flow can be performed, for example, when the real world interface system 200 does not provide a service. Hereinafter, each step of FIG. 10 will be described.

(FIG. 10: Steps S1001 to S1002)
The trajectory observation unit 110 measures the movement trajectory of the autonomous guided vehicle 220 (S1001). The trajectory observation unit 110 calculates the trajectory data 112 of the autonomous guided vehicle 220 based on the photographing result (S1002).

(FIG. 10: Steps S1003 to S1004)
The comparison unit 123 compares the trajectory data 112 with the preset route data 233, and determines the correction amount of the route data 233 (S1003). The route updating unit 124 corrects / updates the route data 233 according to the determined correction amount (S1004).

<Embodiment 2: Supplement>
As a data communication method in the second embodiment, the route update unit 124 is mounted on the autonomous transport vehicle 220, and data is transmitted to and received from the trajectory observation unit 110 through a wireless network. The route update unit 124 is an autonomous transport vehicle. A configuration in which route information is transmitted / received between the route update unit 124 and the autonomous transport vehicle 220 through a wireless network is considered.

<Embodiment 2: Summary>
As described above, in the service system 1000 according to the second embodiment, the trajectory observation unit 110 observes the movement trajectory of the autonomous guided vehicle 220 and updates the route data 233 based on this. Thus, even when the autonomous guided vehicle 220 itself cannot detect that the route data 233 is different from the intended route, it can be corrected based on the objective observation result by the trajectory observation unit 110.

<Embodiment 3>
In Embodiment 3 of the present invention, overall plan control of a transport system using an autonomous transport vehicle in a distribution warehouse or the like will be described. In the transport system, when a general instruction regarding transport is issued, the destination and route of the instructed article are planned by comparing the transport instruction with layout information in the warehouse, current baggage storage status information, and the like. As a method for the transport system to automatically determine this plan, it is possible to determine the transport order, travel route, etc. by solving optimization problems based on travel costs such as transport time and travel distance. Can be considered. A practical algorithm has already been developed, and since it is assumed that this is used in the third embodiment, the details thereof are omitted.

  FIG. 11 is a functional block diagram of the service system 1000 according to the third embodiment. In the third embodiment, the second observation unit 110 is configured as an overall situation observation unit. The first observation unit 210 includes a transport instruction reception unit 212, a layout observation unit 213, and a storage status observation unit 214. The operation instruction unit 230 is configured as a conveyance plan creation / instruction unit. The real world interface system 200 is configured as a transport system for transporting packages. The update instruction unit 120 includes a transport simulator 125 and an algorithm selection / parameter setting unit 126.

  The transport instruction receiving unit 212 receives a transport instruction from the user using a keyboard, a mouse, or the like. The layout observation unit 213 observes the shape of the warehouse and the arrangement of the shelves with a wide range distance sensor or the like, and extracts it as layout information such as a drawing. The storage state observation unit 214 observes how much articles are stored for each shelf using an IC tag or the like, and extracts it as storage state information.

  The transport plan creation / instruction unit 230 creates a plan such as a transport order and a movement route, and controls the operation of the autonomous transport vehicle 220 based on the plan. The transportation plan creation / instruction unit 230 receives each observation information obtained from the first observation unit 210 as an input, creates a transportation plan using the above algorithm using the observation information, and gives the autonomous transportation vehicle 220 a start point, Instruct the end point, loading instructions, route selection, unloading, etc.

  For the method for creating a transportation plan and instructions by the transportation plan creation / instruction unit 230, there are many prepared algorithms for solving the same problem even if the same problem is set. There are several parameters that should be done. Therefore, in order to obtain a more appropriate transport plan, it is important to select an algorithm and set parameters. In general, the operator determines this based on past experience and heuristic considerations.

  If the transport system can always accurately grasp the status of the transported goods and the space availability, it can perform the control as ideal, but generally it cannot be fully understood, so the overall behavior is not necessarily ideal as a result. In addition, it is difficult to confirm whether the algorithm to be used and the parameter settings are optimal. Therefore, in the third embodiment, the algorithm and parameters are changed / corrected based on the actual conveyance status.

  The overall state observation unit 110 includes sensors such as a camera, an IC tag, a distance sensor, and a weight sensor in order to observe the state of the entire indoor such as a target distribution warehouse. The transfer simulator 125 includes information on what algorithms and parameters are selected and usable although the transfer plan creation / instruction unit 230 is currently using or not currently used, and the transfer plan generation / instruction unit 230. The input information such as the conveyance instruction, layout, storage status, etc. grasped by the operator is taken up, and what kind of conveyance can theoretically be performed is simulated. The transfer simulator 125 has a function of creating the same transfer plan and instruction as the transfer plan creation / instruction unit 230 according to the set algorithm and parameters, and how the instruction is issued to the autonomous transfer vehicle 220. Predict whether a moving object will be implemented. Specifically, several simulators that simulate the flow of goods on the physical distribution site have been put into practical use at the commercial level, and it is conceivable that the transport simulator 125 is configured using them. The detailed description is omitted here.

  The algorithm selection / parameter setting unit 126 compares the movement of the object predicted by the transport simulator 125 with the overall situation data 113 obtained by the observation by the overall situation observation unit 110, and evaluates the difference. This difference is the difference between the flow of the object intended by the transport system 200 and the actual flow of the object. The algorithm selection / parameter setting unit 126 changes the algorithm and parameters used by the transport plan creation / instruction unit 230 so as to reduce this difference as much as possible. Specifically, the algorithm and parameters used by the transport simulator 125 are changed variously, and the one with the smallest difference is the current optimal algorithm or parameter, and this is used by the transport plan creation / instruction unit 230. A method for updating the transport system 200 is considered as an example.

  FIG. 12 is a flowchart for explaining a procedure by which the service system 1000 according to the third embodiment updates an algorithm and parameters for creating a transportation plan. Hereinafter, each step of FIG. 12 will be described.

(FIG. 12: Steps S1201 to S1202)
The overall situation observing unit 110 measures the conveyance situation in the entire warehouse or the like (S1201). The transport simulator 125 performs a simulation calculation on what kind of transport can be theoretically performed according to the algorithm and parameters actually used (S1202).

(FIG. 12: Steps S1203 to S1204)
The algorithm selection / parameter setting unit 126 compares the simulation result of the transfer simulator 125 with the actual transfer status observed by the overall status observation unit 110, and determines the algorithm and parameter change method (S1203). The algorithm selection / parameter setting unit 126 changes the algorithm and parameters for transport plan creation actually used by the transport plan creation / instruction unit 230 in accordance with the determined change method (S1204).

<Embodiment 3: Supplement>
In the third embodiment, a configuration in which the transport simulator 125 and the transport plan creation / instruction unit 230 are configured by a fixed computer in a different location from the autonomous transport vehicle 220 can be considered. In this case, it is conceivable that the overall situation observation unit 110, the algorithm selection / parameter setting unit 126, the transport plan creation / instruction unit 230, and the like are connected via a wired or wireless network. It can be considered that they and the autonomous guided vehicle 220 transmit and receive control information and the like via a wireless network.

<Embodiment 3: Summary>
As described above, the service system 1000 according to the third embodiment compares the conveyance status observed by the overall status observation unit 110 and the simulation result by the conveyance simulator 125, and selects the optimal conveyance plan creation algorithm and parameters. be able to.

  The present invention is not limited to the embodiments described above, and includes various modifications. The above embodiment has been described in detail for easy understanding of the present invention, and is not necessarily limited to the one having all the configurations described. A part of the configuration of one embodiment can be replaced with the configuration of another embodiment. The configuration of another embodiment can be added to the configuration of a certain embodiment. Further, with respect to a part of the configuration of each embodiment, another configuration can be added, deleted, or replaced.

  Each of the above-described configurations, functions, processing units, processing means, and the like may be realized in hardware by designing a part or all of them, for example, with an integrated circuit. Each of the above-described configurations, functions, and the like may be realized by software by interpreting and executing a program that realizes each function by the processor. Information such as programs, tables, and files for realizing each function can be stored in a recording device such as a memory, a hard disk, an SSD (Solid State Drive), or a recording medium such as an IC card, an SD card, or a DVD.

  100: Service control system, 110: Second observation unit, 111: Wide range distance sensor data, 112: Trajectory data, 113: Overall situation data, 120: Update instruction unit, 121: Terrain detection unit, 122: Map update unit, 123 : Comparison unit, 124: route update unit, 125: transfer simulator, 126: algorithm selection / parameter setting unit, 200: real world interface system, 210: first observation unit, 211: distance sensor data, 212: transfer instruction reception unit 213: Layout observation unit, 214: Storage state observation unit, 220: Operation unit, 230: Operation instruction unit, 231: Map data, 232: Location estimation unit, 233: Route data, 234: Movement control unit, 1000: Service system.

Claims (7)

A first observation unit for observing a surrounding physical environment; an operation unit for performing a physical operation or information processing based on an observation result of the first observation unit; definition data for defining an operation of the operation unit; and the definition An operation instruction unit for controlling the operation of the operation unit according to data, and a system for controlling a service provided by a real world interface system comprising:
Observation of an observation object different from the first observation part, or a second observation part that observes the same observation object by different means;
When the operating unit is operated on the basis of the observations of the first observation portion updates the definition data based on the observation data obtained by the second observation portion as a result of its operation, the operation unit the first An update unit that updates the definition data based on the observation data obtained by the first observation unit as a result of the operation when operating based on the observation result of the second observation unit in addition to the observation result of the observation unit; ,
Equipped with a,
The definition data is configured as map data describing a map of a place where the real world interface system provides a service, and route data describing a route that the operation unit should move on the map. And service control system. The object observed by the second observation unit includes a physical operation or information processing performed by the operation unit,
The update unit updates the definition data so that a difference between the operation unit operation defined by the definition data and the operation unit operation observed by the second observation unit is small. The service control system according to claim 1. A service control system according to claim 1;
The real world interface system;
A service system characterized by comprising: Before Symbol map data is generated based on the observation data of the second observation portion,
The operation unit is configured as a functional unit in which the autonomous transport vehicle autonomously transports luggage according to the terrain defined by the map data,
The first observation unit is configured using a distance sensor that measures a distance between the autonomous transport vehicle and an object existing around the autonomous transport vehicle.
The second observing unit observes a distance between the second observing unit and the autonomous transport vehicle and an object existing around the second observing unit within a region wider than an observation range of the first observing unit. Configured with sensors,
The operation instruction unit controls a rough movement destination of the autonomous transport vehicle on the terrain based on the observation result of the second observation unit, and the autonomous transport on the map based on the observation result of the first observation unit. Estimating the position and orientation of the car, and based on the result, control the destination more detailed than the approximate destination of the autonomous guided vehicle,
The service system according to claim 3, wherein the updating unit updates the map data so that the observation result of the first observation unit matches the map data. Before SL operating unit, the map data is autonomously guided vehicle according to the terrain and the path defining is configured as a functional unit for autonomously transporting luggage,
The first observation unit is configured using a distance sensor that measures a distance between the autonomous transport vehicle and an object existing around the autonomous transport vehicle.
The second observation unit is configured to observe a trajectory of the autonomous guided vehicle moving on the terrain,
The operation instruction unit estimates the position and orientation of the autonomous transport vehicle on the map based on the observation result of the first observation unit, and based on the result, the autonomous transport vehicle according to the route described by the route data Control the destination of the autonomous guided vehicle so that it moves,
The update unit includes a movement locus of the autonomous transport vehicle obtained from observation data of the second observation unit and route data before the update so that the autonomous transport vehicle moves along the route specified by the route data. The service system according to claim 3, wherein the route data is updated based on the difference. The operation unit is configured as a function unit that autonomously transports a load autonomously,
The definition data is configured as a program that describes an algorithm that defines a procedure for the autonomous transport vehicle to transport the luggage and related parameters,
The first observing unit includes at least one of a layout observing unit that observes a topography of a place where the autonomous transport vehicle transports the luggage, or a storage status observing unit that observes the storage status of the luggage, and the autonomous transport vehicle. It consists of a transport instruction receiving unit that receives transport instructions,
The second observation unit is configured to observe the state of transportation of the luggage by the autonomous transport vehicle,
The operation instruction unit controls an operation of the autonomous transport vehicle transporting the luggage according to the program and the parameter,
The update unit is configured so that the result of simulating the operation of the autonomous transport vehicle transporting the load according to the program and the parameter matches the observation result of the second observation unit. The service system according to claim 3, wherein the service system is updated. The definition data holds a plurality of combinations of the program and the parameter in advance,
The service system according to claim 6, wherein the update unit selects the combination so that the simulation result matches the observation result of the second observation unit.

Images