Mobile body

Abstract

A movable body that moves based on an operational goal, wherein the movable body has: a means for detecting a moving obstacle in the vicinity of the moving body; a means of determining proximity to ascertain whether or not the moving obstacle and the moving body are approaching each other within a predetermined distance; and An operational planning device for planning an evasive maneuver by the moving body and an anticipated evasive maneuver by the moving obstacle, when the proximity determination device determines that the moving obstacle and the moving body are approaching each other within the predetermined distance, further characterized by: a contact tolerance calculation tool for calculating the contact tolerance of the moving body with respect to the moving obstacle, wherein the operation planning tool plans the evasive action of the moving body and the expected evasive action for the moving obstacle on the basis of the contact tolerance calculated by the contact tolerance calculation tool.

Description

Technical field

[0001] The present invention relates to a movable body such as a robot, a manipulator and a mobility device. State of the art

[0002] In the prior art, JP 2008-137 127 A is known in the technical literature in this field. This publication describes a robot control method in which, if it is predicted that an autonomous moving robot and a human will come into contact, the robot's operation is restricted so that it does not come into contact with a vital part of the human. US 2009 / 0 043 440 A1 discloses an autonomous moving body, e.g., a robot, with a sensor device. If the sensor device detects a moving obstacle, it is avoided. For this purpose, motion data from the obstacle and the autonomous body are compared, and the robot's own movement is adjusted. US 2010 / 0 235 033 A1 teaches a moving body, e.g., also a robot. Here, too, an obstacle is detected by the sensor, and an evasive maneuver is initiated.Similarly, US patent 2003 / 0225479A1 describes a method by which a robot can prevent a collision with another object. This involves predicting the movements of moving components in a robot cell and stopping at least one of the components if a collision is imminent. The subsequently published German patent DE112010002021T5 teaches how to detect a vehicle in a blind spot of one's own vehicle by observing that another vehicle has entered the area.

[0003] German patent DE 10 2006 027 187 deals with the driving safety of motor vehicles. It proposes an evasive maneuver when an obstacle in the driving path is detected. To simplify the driver's own evasive maneuver, it can be taken into account that the obstacle also performs a kind of evasive maneuver. EP 2 211 322 A1, which is considered the nearest patent and thus forms the basis for the preamble of claim 1, also teaches the detection of the path of an obstacle, for example, a pedestrian crossing a road. In this case, it is assumed that, if the obstacle is traveling at a sufficiently low speed, it will stop at the edge of the lane and allow the vehicle to pass. Overview of the invention Technical problem

[0004] On the other hand, prior art robot control systems are designed such that, based on the principle of minimizing contact between the robot and humans, the highest priority is given to human operations, and the robot either avoids or restricts these operations. In the prior art robot control system described above, if contact between the robot and humans is predicted, the robot's actions are restricted in accordance with human behavior.

[0005] However, in state-of-the-art control systems where the robot avoids obstacles or restricts the robot's operation in a unilateral manner, a situation continuously arises in busy environments where the robot is likely to come into contact with humans, and the robot may not be able to perform a task efficiently.

[0006] Therefore, it is an object of the invention to provide a movable body that plans an evasive action of the movable body and an expected evasive action that is anticipated by a moving obstacle, thereby efficiently performing a given task. Solution to the task

[0007] To solve the problem described above, the invention provides a movable body that moves along an operational target. The movable body, which moves along the basis of an operational target, includes a means for detecting a movable obstacle in the vicinity of the movable body, a proximity determining means for determining whether or not the movable obstacle and the movable body are approaching each other within a predetermined distance, and an operation planning means for planning, if the proximity determining means determines that the movable obstacle and the movable body are approaching each other within the predetermined distance, an evasive operation by the movable body and an anticipated evasive operation by the movable obstacle.

[0008] According to the movable body of the invention, when a movable obstacle, such as a person, approaches the movable body within a predetermined distance (safety distance or the like), both the movable obstacle and the movable body normally perform an evasive maneuver. Therefore, the anticipated evasive maneuver, which represents an evasive maneuver expected by the movable obstacle, is planned together with the evasive maneuver of the movable body. This allows for an efficient evasive maneuver compared to a case where the movable body evades unlaterally, and enables the efficient execution of a given task.

[0009] It is preferred that the movable body of the invention further includes an alarm operation planning means for planning an alarm operation directed at the movable obstacle on the basis of the expected evasive operation planned by the operation planning means.

[0010] According to the movable body of the invention, the alarm operation that prompts the moving obstacle to perform the anticipated evasive maneuver is carried out, thereby increasing the probability that the moving obstacle understands the intention of the moving body to perform the anticipated evasive maneuver. The moving body then executes the evasive maneuver as expected, allowing for concessions made between the moving body and the moving obstacle that would otherwise be necessary. This enables efficient avoidance or passing with minimal operational effort.

[0011] It is preferred that the movable body of the invention further includes a contact tolerance calculation means for calculating a contact tolerance of the movable body with respect to the movable obstacle, wherein the operation planning means plans the avoidance operation of the movable body and the expected avoidance operation that is expected for the movable obstacle on the basis of the contact tolerance calculated by the contact tolerance calculation means.

[0012] According to the invention's movable body, the evasive maneuver and the anticipated evasive maneuver are planned while allowing a certain degree of contact, thus making it possible to permit a margin of error for the movable body's range of motion compared to a case where no contact is allowed. For this reason, it becomes easier to create a plan so that the movable body can move efficiently even in a busy traffic environment.

[0013] It is preferred that the movable body of the invention further includes a contact tolerance calculation means for calculating a contact tolerance of the movable body with respect to the movable obstacle, wherein the alarm operation planning means plans an alarm operation including a contact operation directed at the movable obstacle on the basis of the contact tolerance calculated by the contact tolerance calculation means.

[0014] According to the movable body of the invention, when the alarm operation, including the contact operation, is planned with respect to the moving obstacle, an alarm, including contact, is directed at a person who has their back turned or the like, thereby making the person aware of the presence of the moving body. A means of applying force or a contact direction is established, allowing the moving body to directly transmit a direction and amount of movement, and an easily transmitted alarm can be executed compared to contactless communication.

[0015] It is preferred that the movable body of the invention further includes a priority relationship determination means for determining the priority relationship between the movable obstacle and the movable body when the proximity determination means determines that the movable obstacle and the movable body are approaching each other within the predetermined distance, wherein the operation planning means plans the avoidance operation of the movable body and the expected avoidance operation of the movable obstacle on the basis of the determination result of the priority relationship determination means.

[0016] According to the movable body of the invention, operations are planned based on the priority relationship between the movable body and the movable obstacle, whereby an appropriate operational plan can be drawn up according to the prevailing circumstances. In particular, if the movable obstacle is a person carrying a heavy load or an elderly person, it is determined that the movable obstacle is given a higher priority, and an operational plan can be drawn up such that the movable body largely avoids it.

[0017] It is preferred that the movable body of the invention further includes a priority relationship determination means for determining the priority relationship between the movable obstacle and the body when the proximity determination means determines that the movable obstacle and the movable body are approaching each other within the predetermined distance, wherein the alarm operation planning means plans the alarm operation directed at the movable obstacle on the basis of the determination result of the priority relationship determination means.

[0018] According to the movable body of the invention, the alarm operation is planned based on the priority relationship between the movable body and the moving obstacle, whereby an appropriate alarm operation can be planned according to the prevailing circumstances. In particular, a weak assertion alarm is executed for a moving object with priority, such as an elderly person, whereby priority can be given to the operation of the moving obstacle. If priority is given to the moving body, a strong assertion alarm is executed, whereby the moving obstacle can take the expected evasive action and priority can be given to the operation of the moving body.

[0019] It is preferred that the movable body of the invention includes a priority relationship determination means for determining the priority relationship between the movable obstacle and the body when the proximity determination means determines that the movable obstacle and the movable body are approaching each other within the predetermined distance, wherein the contact tolerance calculation means calculates the contact tolerance on the basis of the determination result of the priority relationship determination means.

[0020] According to the movable body of the invention, since the contact tolerance can be calculated based on the priority relationship between the movable body and the moving obstacle, contact with a moving obstacle with priority, such as an elderly person, is avoided, thus giving priority to the operation of the moving obstacle. If priority is given to the moving body, contact is positively permitted, thereby giving priority to the operation of the moving body.

[0021] It is preferred that the movable body of the invention further includes an affinity calculation means for calculating, if the movable obstacle is a person, an affinity of the person for the movable body, wherein the operation planning means plans the evasive operation of the movable body and the expected evasive operation expected of the person on the basis of the affinity calculated by the affinity calculation means.

[0022] According to the invention's movable body, operations are planned based on the person's affinity for the movable body, thereby enabling concessions to be made without causing the person discomfort. In particular, if the movable body is an autonomous moving robot, since an elderly person tends to have a low affinity for the robot, the robot will largely avoid the person, thus preventing them from experiencing discomfort.

[0023] It is preferred that the movable body of the invention further includes an affinity calculation means for calculating, if the movable obstacle is a human being, an affinity of the human being for the movable body, wherein the alarm operation planning means plans the alarm operation directed at the movable obstacle on the basis of the affinity calculated by the affinity calculation means.

[0024] According to the movable body of the invention, the alarm operation can be planned based on the person's affinity for the movable body, without causing the person any discomfort. In particular, if the movable body is an autonomous moving robot, since an elderly person tends to have a low affinity for the robot, an alarm operation will not be triggered by contact or strong force, or only a weak alarm operation will be triggered, thus avoiding causing the elderly person discomfort.

[0025] It is preferred that the movable body of the invention further comprises an operation target generation means for generating the operation target of the movable body; and an operation prediction means for predicting the operation of the movable obstacle, which is detected by the means for detecting a movable obstacle, wherein the proximity determination means determines, on the basis of the operation target generated by the operation target generation means and the operation of the movable obstacle predicted by the operation prediction means, whether or not the movable obstacle and the movable body approach each other within the predetermined distance.

[0026] According to the movable body of the invention, the operational goal of the movable body and the predicted operation of the movable obstacle are compared, making it possible to determine with high accuracy whether or not the movable obstacle and the movable body approach each other within the predetermined section.

[0027] In the movable body of the invention, it is preferred that the operation planning means corrects the evasive operation and the expected evasive operation again on the basis of the deviation between the planned evasive operation and the operation target generated by the operation target generation means, and the deviation between the planned expected evasive operation and the operation of the moving obstacle predicted by the operation prediction means.

[0028] According to the movable body of the invention, a correction is performed based on the deviation between the planned evasive operation and the operational goal of the movable body and the deviation between the planned expected evasive operation and the predicted operation of the movable obstacle, thereby enabling the operations to be planned in order to prevent the evasive operation from deviating too much from the operational goal, while preventing the expected evasive operation from imposing an excessive load on the movable obstacle.

[0029] In the movable body of the invention, the operation planning means can plan several operation candidates of the movable body based on the avoidance operation of the movable body and the expected avoidance operation anticipated by the moving obstacle, can predict the operation results of the several operation candidates and select one operation candidate from the several operation candidates for which it is predicted that the operation result will satisfy a predetermined boundary condition.

[0030] According to the movable body of the invention, an operation candidate for which it is predicted that a predetermined boundary condition, such as an approach of the moving obstacle or intensity of an alarm, will be met, is selected from several operation candidates, thereby suppressing an inappropriate operation in practice, such as an excessively intense alarm operation, and enabling an appropriate operation that meets the boundary condition to be performed.

[0031] The movable body according to the invention can be an autonomous movable robot.

[0032] The movable body according to the invention can be a manipulator robot.

[0033] The movable body according to the invention can be a personal mobility device. Advantageous effects of the invention

[0034] According to the invention, the evasive maneuver of the moving body and the expected evasive maneuver of the moving obstacle are planned, thereby enabling the efficient execution of a given task. Brief description of the drawings Fig. Figure 1 is a block diagram representing the configuration of a first embodiment of a movable body according to the invention. Fig. Figure 2 is a flowchart that represents a control procedure of a control device according to the first embodiment. Fig. Figure 3 is a block diagram illustrating the configuration of a second embodiment of a movable body according to the invention. Fig. Figure 4 is a flowchart that represents a control procedure of a control device according to the second embodiment. Fig. Figure 5 is a block diagram illustrating the configuration of a third embodiment of a movable body according to the invention. Fig. Figure 6 is a flowchart that represents a control procedure of a control device according to the third embodiment. Fig. Figure 7 is a block diagram illustrating the configuration of a fourth embodiment of a movable body according to the invention. Fig. Figure 8 is a flowchart that represents a control procedure of a control device according to the fourth embodiment. Fig. Figure 9 is a block diagram illustrating the configuration of a fifth embodiment of a movable body according to the invention. Fig. Figure 10 is a flowchart that represents a control procedure of a control device according to the fifth embodiment. Fig. Figure 11 is a block diagram representing the configuration of a sixth embodiment of a movable body according to the invention. Fig. Figure 12 is a flowchart that represents a control procedure of a control device according to the sixth embodiment. Fig. Figure 13 is a block diagram representing the configuration of a seventh embodiment of a movable body according to the invention. Fig. Figure 14 is a flowchart that represents a control procedure of a control device according to the seventh embodiment. Fig. Figure 15 is a block diagram representing the configuration of an eighth embodiment of a movable body according to the invention. Fig. Figure 16 is a flowchart that represents a control procedure of a control device according to the eighth embodiment. Fig. Figure 17 is a block diagram representing the configuration of a ninth embodiment of a movable body according to the invention. Fig. Figure 18 is a flowchart illustrating a control procedure of a control device according to the ninth embodiment. Fig. Figure 19 is a block diagram representing the configuration of a tenth embodiment of a movable body according to the invention. Fig. Figure 20 is a flowchart that represents a control procedure of a control device according to the tenth embodiment. Description of the embodiments

[0035] Preferred embodiments of the invention are explained in detail below with reference to the drawings. In the drawings, identical or similar sections are designated with the same reference numerals, and overlapping descriptions are not repeated. [First embodiment]

[0036] In a first embodiment, the explanation is provided using an autonomous mobile robot 1 as the mobile body of the invention. The purpose or form of the autonomous mobile robot 1 is not particularly limited. The autonomous mobile robot 1 can be a cleaning robot or a transport robot vehicle. The autonomous mobile robot 1 can move by means of wheels, bipedalism, or the like.

[0037] The autonomous mobile robot 1 plans an evasive maneuver and anticipates a human-directed evasive maneuver if a human operation and a robot operation interfere with each other. It then issues an alarm, prompting the human to execute the anticipated maneuver, thereby realizing concessions made between humans and the robot. According to the autonomous mobile robot 1, its operational goal can be achieved even in busy environments through these concessions with surrounding humans, thus enabling human-robot coexistence.

[0038] This means that people are usually able to move around freely even in very busy environments. However, it is well known that in the event of an accident, crowding (a frozen state) occurs between people everywhere, and orderly movement is important to avoid this crowding. It is reasonable to assume that people move smoothly in a densely packed situation because they are constantly making concessions, even if they are unaware of these concessions. For this reason, concessions made between people are introduced during human-robot operations, thus enabling efficient avoidance.

[0039] It is assumed that concessions between the robot and the human are introduced in three steps: (A) the human understands the robot's intention, (B) the robot understands the human's intention, and (C) the discrepancy between the intentions of the human and the robot is adjusted or reconciled. When implementing concessions, it is important to enable the human to understand the robot's intention. That is, it is preferred that the robot subjectively determines the nature of the relationship when the robot and human interact and how this relationship is presented to the human, and behaves accordingly.

[0040] Accordingly, the autonomous mobile robot 1 plans its own evasive maneuver and the anticipated evasive maneuver expected by the human, and then plans an alarm operation that prompts the human to execute the anticipated evasive maneuver. Examples of alarm operations include nonverbal communication such as gestures, light or sound, direct verbal communication, and the like. In this way, the alarm operation is directed at the human from the autonomous mobile robot 1, thereby transmitting the robot's intention (evasive maneuver or anticipated evasive maneuver) to the human. The human, understanding the intention of the autonomous mobile robot 1, executes the anticipated evasive maneuver, and the robot executes the planned evasive maneuver, thus realizing concessions.

[0041] Next, the configuration of the autonomous mobile robot 1 will be explained.

[0042] As in Fig. As shown in Figure 1, the autonomous moving robot 1 includes a control unit 2 that performs the overall control of the robot. The control unit 2 is formed, for example, by hardware of a microcomputer such as an ECU (electronic control unit) and software. The control unit 2 is connected to a communication unit 3, a position detection unit 4, a map information storage unit 5, a radar sensor 6, and a camera 7.

[0043] Communication unit 3 establishes wireless communication with an external connection to the robot controller. Communication unit 3 receives a task input (operation goal) at the external connection wirelessly. Communication unit 3 then transmits the information of the received task to the controller 2.

[0044] The position acquisition unit 4 acquires the current location of the autonomous mobile robot 1. Position acquisition unit 4 performs position acquisition using GPS (Global Navigation Satellite System for positioning and timing) or a wireless LAN (Local Area Network). During position acquisition, the current location can be obtained by comparing a surrounding obstacle detected by an external sensor, such as a radar sensor or camera, or by comparing the autonomous mobile robot's relative position information with a landmark using map information. Similar to SLAM (Simultaneous Localization and Mapping), a method can be used in which map creation and estimation of the current location by an external sensor are performed simultaneously.The map information storage unit 5 stores a large amount of map information. The position detection unit 4 detects the current location based on the map information in the map information storage unit 5. The position detection unit 4 transmits information about the current location of the autonomous mobile robot 1 and map information in the vicinity of the current location to the control unit 2.

[0045] The radar sensor 6 detects an obstacle in the robot's environment by means of reflected waves of emitted electrical waves. Examples of obstacles include structures such as buildings, people, animals, automobiles, other robots, and the like. A moving obstacle, such as a person, animal, automobile, or other robot, is referred to as a moving obstacle. The radar sensor 6 transmits position information of the detected obstacle to the control unit 2. The camera 7 records the environment of the autonomous moving robot 1. The camera 7 transmits image information of the robot 1's environment to the control unit 2. The radar sensor 6 and the camera 7 function as means for detecting a moving obstacle as described in the claims.Any radar sensor can be used as the radar sensor of the invention, provided that the radar sensor can measure a distance to an obstacle using electromagnetic waves or similar means such as light, electromagnetic waves or sound waves.

[0046] The control unit 2 includes a robot operation target generation unit 10, a human operation prediction unit 11, an operation planning unit 12, a motion control unit 13 and an alarm control unit 14.

[0047] The robot operation target generation unit 10 generates an operation target for the autonomous mobile robot 1. The robot operation target generation unit 10 generates the operation target based on the task information from the communication unit 3, the current location information from the position detection unit 4, and map information. Specifically, if a destination is given as a task, the robot operation target generation unit 10 generates a control target value for an actuator (wheels or the like) for the robot's movement such that the robot arrives at the given destination efficiently. Examples of the target value include a route (position order) or a course (position and time order or velocity order), an acceleration pattern (acceleration order), and the like.The objective of the operation may be to remain in a specific location or to perform a task that is unaccompanied by movement (cleaning under a floor or the like).

[0048] The human operation prediction unit 11 predicts the presence of a human in the vicinity of the autonomous mobile robot 1. The human operation prediction unit 11 first determines whether or not a human is detected in the vicinity of the autonomous mobile robot 1, based on the obstacle's position information from the radar sensor 6 and the image information of the robot's surroundings from the camera 7. The human's position information can be determined using either the obstacle's position information from the radar sensor 6 or the image information of the robot's surroundings from the camera. GPS data of the human can also be obtained through communication. If it is determined that a human is detected, the human operation prediction unit 11 recognizes the human's direction of movement and speed from the history of the surrounding human's position information.The Human Operation Predictor Unit 11 predicts a future operation if the human maintains their direction and speed of movement. Examples of future operations include reducing speed and stopping, changing direction, and remaining in the current position, in addition to continuing in a predetermined direction. The Human Operation Predictor Unit 11 recognizes the appearance of a planned human or the circumstances surrounding the human (positional relationship to an obstacle, a busy environment, or the like), thereby predicting an operation with greater accuracy.

[0049] The operations planning unit 12 plans the operation of the autonomous mobile robot 1. The operations planning unit 12 includes an operations interference determination unit 15, a concession operations planning unit 16, and an alarm operations planning unit 17.

[0050] The operational interference determination unit 15 determines whether or not the surrounding human interferes with the operation of the autonomous mobile robot 1, based on the operational target generated by the robot operational target generation unit 10 and the operation of the surrounding human predicted by the human operation prediction unit 11. In the invention, the operational interference in the operational interference determination unit 15 includes a state in which the human and the autonomous mobile robot 1 approach each other so closely that the safety distance cannot be maintained, as well as a state in which the human and the autonomous mobile robot 1 come into contact with each other. The safety distance is predetermined according to a characteristic, purpose, or the like of the autonomous mobile robot 1.If it is determined that a human and an autonomous mobile robot 1 are approaching each other within the safety distance, the operational interference determination unit 15 determines that the operations interfere with each other. The operational interference determination unit 15 functions as a proximity determination device, which is explained in the claims. The safety distance of the autonomous mobile robot 1 corresponds to a predetermined distance, which is explained in the claims.

[0051] If the Operational Interference Determination Unit 15 determines that a nearby human is interfering with the operation of the autonomous mobile robot 1, the Concessional Operation Planning Unit 16 plans a concessional operation. The concessional operation is an operation that includes the avoidance operation of the autonomous mobile robot 1 and an anticipated avoidance operation that is expected by the human interfering with the operation of the autonomous mobile robot. The avoidance operation is the operation of the autonomous mobile robot 1 to avoid interference with the human. The anticipated avoidance operation is an avoidance operation that can be expected by the human to avoid approaching or contacting the autonomous mobile robot 1 as much as possible. The Concessional Operation Planning Unit 16 can plan the anticipated avoidance operation using historical statistical data or the like.

[0052] The concession operation planning unit 16 plans the evasive action that enables the most efficient avoidance while ensuring a safe distance, based on the planned expected evasive action and the operation target generated by the robot operation target generation unit 10. The concession operation planning unit 16 can plan the evasive action of the autonomous mobile robot 1 and can then plan the expected evasive action of the human based on the planned evasive action.

[0053] The concession operation planning unit 16 corrects the evasive operation and the expected evasive operation based on the deviation between the planned evasive operation and the operational target generated by the robotic operation target generation unit 10, and the deviation between the planned expected evasive operation and the human operation predicted by the human operation prediction unit 11. For example, the concession operation planning unit 16 makes corrections to minimize the two deviations. Therefore, the concession operation planning unit 16 can correct the plan to prevent the evasive operation from deviating from the operational target while preventing the expected evasive operation from negatively impacting the human. The concession operation planning unit 16 can correct the plan to minimize the sum of the two deviations.The concession operations planning unit 16 functions as an operations planning tool, which is explained in the claims.

[0054] The Alarm Operations Planning Unit 17 plans an alarm operation that prompts the person to perform the anticipated evasive action, based on the anticipated evasive action planned by the Concession Operations Planning Unit 16. Alarm Operations Planning Unit 17 plans the alarm operation using communication methods such as verbal communication, gestures, light, or sound. In the case of a verbal alarm, the direction and extent of the evasive action can be communicated verbally.In the case of a gesture-based alarm, for example, a movable component such as a robot arm is used. The direction of the person's avoidance can be represented by the direction of movement of the movable component or the direction of a cue section of the movable component, and the degree of avoidance can be represented by the direction of movement of the movable component or the size of the area of ​​movement. In the case of a light-based alarm, for example, several light sources can be arranged in parallel to emit light sequentially. The direction of avoidance can be represented by the direction in which the light flows and is visible, and the degree of avoidance can be represented by the intensity of the light. In the case of a sound-based alarm, for example, the degree of avoidance can be conveyed by the volume of the sound.These alarm procedures can be used in combination. The explanation will not be repeated in the following sections.

[0055] The motion control unit 13 controls a movement mechanism of the autonomous mobile robot 1. The motion control unit 13 controls a movement mechanism based on the evasion operation planned by the concession operation planning unit 16 to execute the evasion operation of the autonomous mobile robot 1. The motion control unit 13 controls the movement mechanism along the operation target generated by the robot operation target generation unit 10, thereby achieving the operation target.

[0056] The alarm control unit 14 controls a speech output unit, a light output unit, or a common drive attachment of the autonomous moving robot 1. The alarm control unit 14 controls the speech output section, the light output unit, or the common drive section based on the alarm operation planned by the alarm operation planning unit 17 to execute the alarm operation using speech, light, or a gesture through the position or movement of an upper body section.

[0057] Next, a tax procedure will be provided in the tax unit 2 described above.

[0058] As in Fig. As shown in Figure 2, the robot operation target generation unit 10 in the control unit 2 first performs an operation target generation process to generate the operation target of the autonomous mobile robot (S1). The robot operation target generation unit 10 generates the operation target based on the task information from the communication unit 3, the information about the current location from the position detection unit 4, and the map information.

[0059] Next, the human operation prediction unit 11 determines whether or not a human is detected in the vicinity of the autonomous moving robot 1, based on the position information of the obstacle from the radar sensor 6 and the image information of the robot's surroundings from the camera 7 (S2). If it is determined that no human is detected in the vicinity of the autonomous moving robot, the human operation prediction unit 11 proceeds to step S5. The operation of the invention is not limited to humans and the invention can be applied to all moving obstacles that move autonomously, such as robots and automobiles.

[0060] If it is determined that a person is detected in the vicinity of the autonomous mobile robot 1, the human operation prediction unit 11 performs operation prediction processing to predict the operation of the detected surrounding person (S3). The human operation prediction unit 11 detects the direction and speed of movement of the surrounding person based on the position information of the obstacle from the radar detection direction 6 and the image information in the robot's environment from the camera 7. The human operation prediction unit 11 predicts a future operation by the person if the direction and speed of movement are maintained.

[0061] The operation interference determination unit 15 determines whether the surrounding human and the operation of the autonomous mobile robot 1 interfere with each other, based on the operation target generated by the robot operation target generation unit 10 and the operation of the surrounding human predicted by the human operation prediction unit 11 (S4).

[0062] If it is determined that no interference is present, the operational interference determination unit 15 continues with step S5.

[0063] In step S5, the motion control unit 13 performs normal robot control processing to control the motion mechanism of the autonomous moving robot 1 along the operation target generated by the robot operation target generation unit 10.

[0064] In step S4, if the Operational Interference Determination Unit 15 determines that no operational interference exists, the Concessional Operation Planning Unit 16 executes a concessional operation planning process to plan a concessional operation (S6). The concessional operation planning unit 16 plans an evasive operation by the autonomous mobile robot 1 and an expected evasive operation by a human as the concessional operation. Subsequently, the Alarm Operation Planning Unit 17 executes an alarm operation planning process to plan an alarm operation for the human based on the expected evasive operation planned by the concessional operation planning unit 16 (S7).

[0065] In step S8, the motion control unit 13 and the alarm control unit 14 execute interference-assisted robot control processing to implement the concession operation. In this interference-assisted robot control processing, the motion control unit 13 controls the movement mechanism based on the avoidance operation of the autonomous moving robot 1, which is planned by the concession operation planning unit 16. The alarm control unit 14 controls the autonomous moving robot 1 so that it executes the alarm operation planned by the alarm operation planning unit 17.

[0066] The control unit 2 then determines whether the operational goal has been achieved and the operation of the autonomous mobile robot 1 ends (S9). If it is determined that the operation of the autonomous mobile robot 1 does not end, the control unit 2 repeats the processing starting from step S1. If it is determined that the operation of the autonomous mobile robot 1 ends, the control unit 2 terminates the control process.

[0067] The functional effects of the autonomous mobile robot 1 described above will then be explained.

[0068] According to the autonomous mobile robot 1 of the first embodiment, when a person approaches the autonomous mobile robot 1 within the safety distance, both the person and the autonomous mobile robot 1 execute the evasive maneuver normally. Therefore, the anticipated evasive maneuver, which represents the maneuver expected by the person, is planned together with the evasive maneuver of the autonomous mobile robot 1. This allows for an efficient evasive maneuver compared to a case where the autonomous mobile robot 1 performs the evasive maneuver unilaterally, and a given task can be performed efficiently. Consequently, movement efficiency is improved, enabling the autonomous mobile robot 1 to perform a task in a short time, and the operational efficiency of the autonomous mobile robot 1 is enhanced.

[0069] According to the autonomous mobile robot 1, an alarm operation is executed, prompting the human to perform the anticipated evasive maneuver. This increases the likelihood that the human will understand the robot's intention and execute the anticipated maneuver. The autonomous mobile robot 1 then performs the evasive maneuver as expected, thus enabling concessions made between humans and the autonomous mobile robot 1. Therefore, according to the autonomous mobile robot 1, the operational objective can be achieved even in busy environments by making concessions to surrounding humans, and human-robot coexistence can be realized.

[0070] The intention of the autonomous mobile robot 1 is transmitted to an approaching person, thus eliminating the need for the person to give the autonomous mobile robot 1 a greater distance than necessary. This allows the person and the autonomous mobile robot 1 to pass each other efficiently in a busy environment. The intention of the autonomous mobile robot 1 is transmitted via an alarm function, preventing any incorrect maneuvers and collisions with the autonomous mobile robot 1. This allows the person to experience a sense of ease and freedom of movement. [Second embodiment]

[0071] As in Fig. As shown in Figure 3, an autonomous mobile robot 21 of a second embodiment differs from the autonomous mobile robot 1 of the first embodiment in that a priority relationship determination unit 24, which determines the priority relationship between the human and the autonomous mobile robot 21, is provided, and the concession operation and the alarm operation are planned on the basis of the priority relationship determined by the priority relationship determination unit 24.

[0072] In the autonomous mobile robot 21 of the second embodiment, the priority relationship between the interfering humans and the autonomous mobile robot 21 is determined. If the degree to which a new motion goal planned for the evasive operation can maintain an original motion goal is defined as motion efficiency in the concession operation, the motion efficiencies of the human and the autonomous mobile robot 21 exhibit a so-called compromise relationship. The priority relationship is used as an index for resolving the compromise. For example, the new motion goals of the human and the autonomous mobile robot 21 are determined such that the ratio of the motion efficiencies of the human and the autonomous mobile robot 21 corresponds to the priority relationship.If the human has high priority, the original movement goal of the human is maintained to a high percentage compared to the original movement goal of the autonomous moving robot 21.

[0073] In the autonomous mobile robot 21, the priority relationship determination unit 24 of the control device 22 determines the priority relationship between the human and the autonomous mobile robot 21. The priority relationship determination unit 24 compares the state of the autonomous mobile robot 21 and the state of the human to determine how much priority should be given to either the autonomous mobile robot 21 or the human.

[0074] The priority relationship determination unit 24 detects a state in which a person is in a hurry or is carrying a large or heavy load, based on the obstacle's position information from the radar sensor 6 and the image information from the camera 7. Each state, using a suitable sensor, can be identified by a value obtained by measuring or estimating a physical quantity representing the intensity of movement, such as the amount of movement, mass, or kinetic energy. The priority relationship determination unit 24 determines that the person in a hurry or carrying a large or heavy load has high priority. The priority relationship determination unit 24 also detects the person's type, such as age or gender, and determines that a woman, a child, or an elderly person has high priority.The content of a task to be performed or the condition of a person can be identified, and it can be determined that a person who is in a wheelchair, a leader in a crowd, or the like has high priority.

[0075] The priority relationship determination unit 24 determines the priority of the autonomous mobile robot 21 according to the importance or urgency of the operational objective. The priority relationship determination unit 24 determines the importance or urgency of the operational objective based on its content, time constraints, or similar factors. For example, if the content of the operational objective relates to human safety, such as saving lives, or if the operational objective has a deadline, the robot is determined to have high priority. The importance or urgency of the operational objective can be manually set by a human when assigning the operational objective. If the importance or urgency of the operational objective is high, the priority relationship determination unit 24 determines that the autonomous mobile robot 21 has high priority.Similar to determining the priority with respect to humans, the priority relationship determination unit 24 can determine that the priority is high if the movement speed of the autonomous mobile robot 21 is high or the transport load is large, that is, if a physical quantity representing the intensity of the movement is large.

[0076] The concession operation planning unit 25 plans the evasive action of the autonomous mobile robot 21 and the anticipated evasive action of the human based on the determination result of the priority relationship determination unit 24, the operation goal generated by the robot operation goal generation unit 10, and the human operation predicted by the human operation prediction unit 11. If the human has high priority, the concession operation planning unit 25 maintains the original movement goal as much as possible and plans the anticipated evasive action close to the human operation predicted by the human operation prediction unit. That is, the anticipated evasive action is planned in such a way that the human, who has high priority, is not subjected to excessive strain.If the autonomous mobile robot 21 has a high priority, the concession operation planning unit 25 plans the evasion operation, in which the original motion goal is maintained as much as possible to achieve the operation goal on a priority basis. It is preferred that the extent to which the evasion operation maintains the motion goal is quantified from the amount of change in the direction of travel or the speed of the evasion operation for the motion goal, or from an increase in the required travel time or distance to a given point, or from energy consumption.

[0077] The alarm operation planning unit 26 plans the alarm operation based on the determination result of the priority relationship determination unit 24 and the expected evasive action planned by the concession operation planning unit 25. If the human has high priority, the alarm operation planning unit 26 plans an alarm operation with weak assertion. If the autonomous mobile robot 21 has high priority, the alarm operation planning unit 26 plans an alarm operation with strong assertion. In the invention, the setting or adjustment of the assertion intensity of the alarm operation is achieved by the volume of a sound in the case of an alarm triggered by sound, the brightness of light in the case of an alarm triggered by light, and the magnitude of a force at the time of contact in the case of an alarm triggered by contact. The explanation is not repeated here for the same case.

[0078] Next, a control procedure of the control device 22 of the second embodiment will be explained.

[0079] As in Fig. As shown in Figure 4, the robot operation target generation unit 10, controlled by the control unit 22, first performs an operation target generation process to generate the operation target (S11). Next, the human operation prediction unit 11 determines whether or not a person is detected in the environment, based on the position information of the obstacle for the radar sensor 6 and the image information of the robot's environment from the camera 7 (S12). If it is determined that no person is detected, the human operation prediction unit 11 proceeds to step S15.

[0080] If it is determined that a human is being detected, the human operation prediction unit 11 performs operation prediction processing to predict the operation of the detected human (S13). Subsequently, the operation interference determination unit 15 determines whether or not the surrounding person interferes with the operation of the autonomous mobile robot 21, based on the operation target generated by the robot operation target generation unit and the human operation predicted by the human operation prediction unit 11 (S14). If it is determined that no operation interference is present, the operation interference determination unit 15 proceeds to step S15.

[0081] In step S15, the motion control unit 13 performs normal robot control processing to control the motion mechanism of the autonomous moving robot 21 along the operation target generated by the robot operation target generation unit 10.

[0082] In step S14, when the operation interference determination unit 15 determines that an operation interference is present, the priority relationship determination unit 24 executes a priority relationship determination processing to determine the priority relationship between the autonomous mobile robot 21 and the human (S16).

[0083] Subsequently, the concession operation planning unit 25 performs concession operation planning processing based on the determination result of the priority relationship determination unit 24, the operation goal generated by the robot operation goal generation unit 10, and the human operation predicted by the human operation prediction unit 11 (S17). The concession operation planning unit 25 plans the evasive action of the autonomous mobile robot 21 and the expected evasive action anticipated by the human as the concession operation. Following this, the alarm operation planning unit 26 performs alarm operation planning processing to plan the alarm operation directed at the human based on the determination result of the priority relationship determination unit 24 and the expected evasive action planned by the concession operation planning unit 25 (S18).

[0084] In step S19, the motion control unit 13 and the alarm control unit 14 perform interference-assisted robot control processing to implement the concession operation. In this interference-assisted robot control processing, the motion control unit 13 controls the movement mechanism based on the avoidance operation of the autonomous moving robot 21, which is planned by the concession operation planning unit 25. The alarm control unit 14 controls the autonomous moving robot 21 to execute the alarm operation, which is planned by the alarm operation planning unit 26.

[0085] The control unit 22 determines whether the operational goal has been achieved and whether the operation of the autonomous moving robot 21 ends (S20). If it is determined that the operation of the autonomous moving robot 21 does not end, the control unit 22 repeats the processing starting from step S11. If it is determined that the operation of the autonomous moving robot 21 ends, the control unit 22 terminates the control.

[0086] According to the autonomous mobile robot 21 of the second embodiment, operations are planned based on the priority relationship between the autonomous mobile robot 21 and the human, whereby an appropriate operational plan can be created according to the circumstances. In particular, if the other party is a human carrying a heavy load or an elderly person, it is determined that the human has high priority, and an operational plan such as the autonomous mobile robot 21 largely performing evasive maneuvers can be created. If the autonomous mobile robot 21 is moving with great urgency, such as for medical treatment, it is determined that the autonomous mobile robot 21 has high priority, and an operational plan in which the human largely performs evasive maneuvers can be created.

[0087] According to the autonomous mobile robot 21, the alarm operation can be planned based on the priority relationship between the autonomous mobile robot 21 and the human. A low-priority alarm operation is performed for an elderly person with high priority, in which case the human operation can be given high priority. If the autonomous mobile robot 21 has high priority, a high-priority alarm operation is performed, allowing the human to take the expected evasive action, and the autonomous mobile robot 21 operation can be given priority. [Third embodiment]

[0088] As in Fig. Figure 5 shows that an autonomous mobile robot 31 of a third embodiment differs from the autonomous mobile robot 1 of the first embodiment in that contact between the human and the autonomous mobile robot 31 is permitted, a contact sensor 33 is provided, a contact tolerance calculation unit 35 is provided which calculates a contact tolerance between the human and the autonomous mobile robot 31, and the concession operation and the alarm operation are planned on the basis of the contact tolerance calculated by the contact tolerance calculation unit 35.

[0089] The autonomous mobile robot 31 of the third embodiment creates an operational plan while allowing contact between the human and the autonomous mobile robot 31. In this case, a plan is created while contact is permitted in advance and the safety distance is reduced, thereby decreasing the effective traffic density to allow for route flexibility. However, it is necessary to ensure human safety even when contact is established. The distance is set to 0 or a negative value, allowing a plan to be created such that the autonomous mobile robot 31 makes positive contact with the human. Consequently, the route has a large margin of safety, making it easy to create a plan that allows the autonomous mobile robot 31 to move efficiently.

[0090] In the autonomous mobile robot 31, contact is used as a positive form of communication with humans during alarm operations. For example, contact with a person who has their back turned can make the person aware of the autonomous mobile robot 31's presence. A path for exerting force or a direction of contact is set or adjusted, whereby the "direction" and "amount" of movement can be directly transmitted, and an easily transmitted alarm can be executed in a simpler manner compared to contactless communication.

[0091] The contact sensor 33 of the third embodiment is, for example, a pressure sensor or a force sensor or the like, and detects a contact state such as the presence / absence of contact between the obstacle and the autonomous moving robot 31, the intensity of a contact pressure, or a contact direction. The contact sensor 33 transmits the information about the detected contact state to the control unit 32.

[0092] The contact tolerance calculation unit 35 calculates contact tolerance based on the obstacle's position information from the radar sensor 6, the image information from the camera 7, and the contact status information from the contact sensor 33. Specifically, the contact tolerance calculation unit 35 detects the orientation of a person's face or body using image processing and calculates the contact tolerance accordingly. Since a person not facing the direction of the autonomous moving robot 31 would have difficulty being aware of its presence, the contact tolerance is calculated to be high, allowing the person to become aware of the robot's presence through contact.Alternatively, to avoid causing discomfort to a person unaware of the autonomous moving robot 31 due to abrupt contact, the contact tolerance can be calculated to be low for a person not facing the direction of the autonomous moving robot 31, thus avoiding contact as much as possible. It is preferred that an appropriate calculation method for the contact tolerance be selected according to circumstances that take into account the balance of the two approaches described above.

[0093] The concession operation planning unit 36 ​​plans the evasive maneuver of the autonomous mobile robot 31 and the expected evasive maneuver of the human based on the contact tolerance calculated by the contact tolerance calculation unit 35, the operational goal generated by the robot operational goal generation unit 10, and the human operation predicted by the human operation prediction unit 11. The concession operation planning unit 36 ​​can reduce the safety distance specified for the human in a route because the contact tolerance is high. This means that the effective physical size of a human can be reduced, and the route can have a margin of error, thus enabling a more efficient evasive maneuver to be planned.

[0094] The alarm operation planning unit 37 plans the alarm operation based on the contact tolerance calculated by the contact tolerance calculation unit 35 and the expected operation planned by the concession operation planning unit 36. The alarm operation planning unit 37 plans the alarm operation by contact based on the obstacle position information from the radar sensor 6, the image information from the camera 7, and the contact status information from the contact sensor 33, if it is determined that the alarm operation by contact is valid. In the case of an alarm by contact, the direction of evasion of the person can be determined by the direction of the force at the time of contact, and the degree of evasion can be determined by the magnitude of this force. The description for the same case will not be repeated hereafter.The alarm operation planning unit 37 plans the alarm operation using direct communication via contact, whereby the intention of the autonomous mobile robot 31 can be transmitted to the human more reliably. Next, a control procedure of the control unit 32 of the third embodiment is explained.

[0095] As in Fig. As shown in Figure 6, the robot operation target generation unit 10 in the control unit 32 first performs an operation target generation process to generate the operation target (S31). Next, the human operation prediction unit 11 determines whether or not a human is detected in the environment, based on the position information of the obstacle from the radar 6, the image information of the robot's environment from the camera 7, and the contact status information from the contact sensor 33 (S32). If it is determined that no human is detected, the human operation prediction unit 11 proceeds to step S35.

[0096] If it is determined that a human is being detected, the human operation prediction unit 11 performs operation prediction processing to predict the operation of the detected human (S33). Subsequently, the operation interference determination unit 15 determines whether or not the surrounding person interferes with the operation of the autonomous mobile robot 31, based on the operation target of the robot operation target generation unit 10 and the human operation predicted by the human operation prediction unit 11 (S34). If it is determined that no operation interference is present, the operation interference determination unit 15 proceeds to step S35.

[0097] In step S35, the motion control unit performs robot control processing to control the motion mechanism of the autonomous mobile robot 31 along the motion target generated by the robot operation target generation unit 10.

[0098] In step S34, when the Operational Interference Determination Unit 15 determines that operational interference is present, the Contact Tolerance Calculation Unit 35 performs a contact tolerance calculation processing to calculate the contact tolerance based on the position information of the obstacle from the radar sensor 6, the image information from the camera 7 and the contact state information from the contact sensor 33 (S36).

[0099] Subsequently, the concession operation planning unit 36 ​​performs concession operation planning processing based on the contact tolerance calculated by the contact tolerance calculation unit 35, the operation goal generated by the robot operation goal generation unit 10, and the human operation predicted by the human operation prediction unit 11 (S37). The concession operation planning unit 25 plans the avoidance operation of the autonomous mobile robot 21 and the expected avoidance operation anticipated by the human operating the autonomous mobile robot 31 as the concession operation.Subsequently, the Alarm Operations Planning Unit 37 performs an Alarm Operations Planning process to plan the alarm operation for the human based on the contact tolerance calculated by the Contact Tolerance Calculation Unit 35 and the expected avoidance operation planned by the Concession Operations Planning Unit 36 ​​(S38).

[0100] In step S39, the motion control unit 13 and the alarm control unit 14 perform interference-assisted robot control processing to implement the concession operation. In this interference-assisted robot control processing, the motion control unit 13 controls a movement mechanism based on the avoidance operation planned by the concession operation planning unit 36 ​​to execute the avoidance operation of the autonomous moving robot 31. The alarm control unit 14 controls the autonomous moving robot 31 to execute the alarm operation planned by the alarm operation planning unit 37.

[0101] The control unit 32 then determines whether the operational goal is achieved and the operation of the autonomous moving robot 31 ends (S40). If it is determined that the operation of the autonomous moving robot 31 does not end, the control unit 32 repeats the processing of step S31. If it is determined that the operation of the autonomous moving robot 31 ends, the control unit 32 terminates the control process.

[0102] According to the autonomous mobile robot 31 of the third embodiment, the evasive maneuver and the anticipated evasive maneuver are planned while allowing a certain degree of contact, making it possible to allow a margin for the autonomous mobile robot 31's range of motion compared to a case where no contact is permitted. For this reason, it is possible to allow a margin for the route even in a busy environment, and it becomes easy to create a plan such that the autonomous mobile robot 31 can move efficiently.

[0103] According to the autonomous mobile robot 31, it will be possible to traverse the route even when contact is unavoidable. Since contact is used as a form of positive communication with humans during alarm operations, an easily transmitted alarm can be executed more efficiently compared to contactless communication, and the intention of the autonomous mobile robot 31 can be transmitted to the human more reliably. [Fourth embodiment]

[0104] As in Fig. As shown in Figure 7, an autonomous mobile robot 41 of a fourth embodiment differs from the autonomous mobile robot 31 of the third embodiment in that a priority relationship determination unit 44, which determines the priority relationship between the human and the autonomous mobile robot 41, is provided, contact tolerance is calculated on the basis of the priority relationship calculated by the priority relationship determination unit 44, and the concession operation and the alarm operation are planned on the basis of the priority relationship determined by the priority relationship determination unit 44.

[0105] In the autonomous mobile robot 41 of the fourth embodiment, the concession operation and the alarm operation are planned using the priority relationship between the human and the autonomous mobile robot 41 and the contact tolerance of the autonomous mobile robot 41 with respect to the human.

[0106] The priority relationship determination unit 44 of the control device 42 of the fourth embodiment determines the priority relationship between the human interfering with the operation of the autonomous moving robot 41 and the autonomous moving robot 41 based on the contact state information of the contact sensor 33 in addition to the position information of the obstacle from the radar sensor 6 and the image information from the camera 7. The priority relationship determination unit 44 uses the contact state information of the contact sensor 33, whereby the state in which it is difficult for the human to perform evasive maneuvers due to a heavy load or the like can be recognized, and the priority relationship can be appropriately determined.

[0107] The contact tolerance calculation unit 45 calculates the contact tolerance based on the determination result of the priority relationship determination unit, in addition to the obstacle position information from radar sensor 6, the image information from camera 7, and the contact state information from contact sensor 33. Specifically, the contact tolerance calculation unit 35 calculates the contact tolerance as low because it is not desirable to establish contact with a person carrying a heavy load and having high priority, or with a person who does not permit contact. The contact tolerance calculation unit 35 calculates the contact tolerance as high when the autonomous mobile robot 41 has high priority, allowing for some leeway in its route and ensuring efficient movement. The type of contact is considered in terms of contact capability or degree.The adjustment of the type of contact establishment can be carried out by changing the movement relationship between the human and the autonomous moving robot 41.

[0108] The concession operation planning unit 46 plans the evasive action of the autonomous mobile robot 41 and the expected evasive action of the human based on the contact tolerance calculated by the contact tolerance calculation unit 45, the operation goal generated by the robot operation goal generation unit 10, the human operation predicted by the human operation prediction unit 11, and the determination result of the priority relationship determination unit 44. Similarly, the alarm operation planning unit 47 plans the alarm operation based on the contact tolerance calculated by the contact tolerance calculation unit 45, the expected evasive action planned by the concession operation planning unit 46, and the determination result of the priority relationship determination unit 44.

[0109] The alarm operation planning unit 37 plans the alarm operation by contact based on the position information of the obstacle from the radar sensor 6, the image information from the camera 7, and the contact status information from the contact sensor 33, provided the alarm operation by contact is valid. In this case, the alarm operation is planned using direct communication by contact, which allows the intention of the autonomous moving robot 41 to be transmitted more reliably to the human. Next, a control procedure of the control unit 42 of the fourth embodiment is explained.

[0110] As in Fig. As shown in Figure 8, the robot operation target generation unit 10 in the control unit 42 first performs an operation target generation process to generate the operation target (S51). Next, the human operation prediction unit 11 determines whether or not a human is detected in the environment, based on the position information of the obstacle from the radar sensor 6, the image information of the robot's environment from the camera 7, and the contact status information from the contact sensor 33 (S52). If it is determined that no human is detected, the human operation prediction unit 11 proceeds to step S55.

[0111] If it is determined that a human is being detected, the human operation prediction unit 11 performs operation prediction processing to predict the operation of the detected human (S53). Subsequently, the operation interference determination unit 15 determines whether or not the surrounding person interferes with the operation of the autonomous mobile robot 41, based on the operation target from the robot operation target generation unit 10 and the human operation predicted by the human operation prediction unit 11 (S54). If it is determined that no operation interference is present, the operation interference determination unit 15 proceeds to step S55.

[0112] In step S55, the motion control unit 13 performs normal robot control processing to control the motion mechanism of the autonomous moving robot 41 along the operation target generated by the robot operation target generation unit 10.

[0113] In step S54, when the Operational Interference Determination Unit 15 determines that operational interference exists, the Priority Relationship Determination Unit 44 performs priority relationship determination processing to determine the priority relationship between the autonomous moving robot 41 and the human (S56). Subsequently, the Contact Tolerance Calculation Unit 45 performs contact tolerance calculation processing to calculate the contact tolerance based on the obstacle position information from the radar sensor 6, the image information from the camera 7, and the contact state information from the contact sensor 33 (S57).

[0114] Subsequently, the concession operation planning unit 46 performs concession operation planning processing based on the determination result of the priority relationship determination unit 44, the contact tolerance calculated by the contact tolerance calculation unit 45, the operation goal generated by the robot operation goal generation unit 10, and the human operation predicted by the human operation prediction unit 11 (S58). The concession operation planning unit 46 plans, as the concession operation, the avoidance operation of the autonomous mobile robot 41 and the expected avoidance operation of the human interfering with the operation of the autonomous mobile robot 41.Subsequently, the Alarm Operations Planning Unit 47 performs an Alarm Operations Planning process to plan the alarm operation for the human on the basis of the contact tolerance calculated by the Contact Tolerance Calculation Unit 45 and on the basis of the expected avoidance operation planned by the Concession Operations Planning Unit 46 (S59).

[0115] In step S60, the motion control unit 13 and the alarm control unit 14 execute interference-assisted robot control processing to implement the concession operation. In this interference-assisted robot control processing, the motion control unit 13 controls the movement mechanism based on the avoidance operation of the autonomous moving robot 41, which is planned by the concession operation planning unit 46. The alarm control unit 14 controls the autonomous moving robot 41 so that it executes the alarm operation, which is planned by the alarm operation planning unit 47.

[0116] The control unit 42 then determines whether the operational goal is achieved and the operation of the autonomous moving robot 41 ends (S61). If it is determined that the operation of the autonomous moving robot 41 does not end, the control unit 42 repeats the processing of step S51. If it is determined that the operation of the autonomous moving robot 41 ends, the control unit 42 terminates the control process.

[0117] According to the autonomous moving robot 41 of the fourth embodiment, the same effects can be achieved as with the autonomous moving robots 21 and 31 of the second and third embodiments, respectively. The priority relationship determination unit 44 determines the priority relationship by also taking the contact state information into account, thus enabling a more reliable determination. The contact tolerance calculation unit 45 calculates the contact tolerance by taking the priority relationship into account, thus enabling the contact tolerance to be calculated according to the circumstances and allowing an efficient operating plan to be created. [Fifth embodiment]

[0118] As in Fig. As shown in Figure 9, an autonomous mobile robot 51 of a fifth embodiment differs from the autonomous mobile robot 1 of the first embodiment in that contact between the human and the autonomous mobile robot 51 is allowed, an affinity calculation unit 54, which calculates an affinity of the human for the autonomous mobile robot 51, is provided, and the concession operation and the alarm operation are planned on the basis of the affinity calculated by the affinity calculation unit 54.

[0119] In the autonomous mobile robot 51 of the fifth embodiment, a human affinity for the autonomous mobile robot 51 is calculated. Affinity signifies a friendly disposition between the robot and the human from physical, informal, emotional, environmental, and economic perspectives, and represents a human's feeling of fear or ease towards the robot. It is assumed that the greatest difference between concessions between humans and concessions between humans and robots is the human's affinity for the robot. Affinity depends on the value placed on an individual or the impact of the robot and becomes an evaluation criterion when a behavior triggers an alarm. Children who are strongly unsettled by the robot and people who are afraid of machines will change their behavior or way of thinking towards the robot.In the case of children, they simply approach the robot, which, conversely, impairs the robot's movement. However, from the perspective of ease, no discomfort is conveyed, even if the distance is relatively small and there is some leeway in route selection. In cases where people are afraid, they will move away from the robot, thus ensuring the robot's route is followed. This can cause discomfort for the person, even if the robot is at a distance and the route options are limited. Consequently, the movement relationship between the person and the robot is adjusted or aligned according to the affinity for modifying the robot's operation, thereby creating an operational plan without causing discomfort for the person.Adjusting the movement relationship according to affinity can be achieved, for example, by increasing the distance to the human when the affinity is low, or by reducing the robot's movement speed when it approaches the human, or similar measures.

[0120] The control unit 52 of the autonomous mobile robot 51 calculates the human's affinity for the autonomous mobile robot 51 in the affinity calculation unit 54. The affinity calculation unit 54 calculates the affinity based on the image information from the camera 7.

[0121] In particular, the affinity calculation unit 54 first recognizes the size, weight, and type, such as age or sex, based on the image information from the camera 7. The affinity calculation unit 54 stores parameters relating to the size, weight, or color of the autonomous moving robot 51 and a contact risk of the robot's surface in advance. The affinity calculation unit 54 also calculates statistical data regarding the human's affinity for the autonomous moving robot 51. It is preferred that the statistical data regarding affinity include data regarding the distance and physiological state (heart rate, blood pressure, or the like) when the human approaches the autonomous moving robot 51. In this case, it is assumed that if the human's heart rate or blood pressure increases at a distance, the affinity is low.

[0122] The affinity calculation unit 54 calculates the affinity based on a comparison of the size of a person and the size of the autonomous mobile robot 51, a comparison of the weight of a person and the weight of the autonomous mobile robot 51, the type of person (e.g., gender or age), the color of the autonomous mobile robot 51, the contact risk of the surface, and statistical data. The affinity calculation unit 54 calculates the affinity as low if the size of the autonomous mobile robot is larger than the size of a person. Parameters relating to size include height, width, area, volume, or similar dimensions.For example, starting from the autonomous moving robot 41, the area of ​​the human in the image of camera 7 is considered the projection area, and in this case, from the perspective of the human as the other party, the size of the autonomous moving robot is calculated as the projection area, thus making the comparison of size appropriate.

[0123] Affinity calculation unit 54 calculates the affinity as low if the weight of the autonomous moving robot 51 is greater than the weight of the human. Affinity calculation unit 54 calculates the affinity as low if the contact risk (edge ​​sharpness or the like) of the robot's surface is high. Affinity calculation unit 54 calculates the affinity as low if the luminosity of the robot's color is low, if the saturation is high, or if the color is associated with risk or increased alertness (red, yellow, or the like). Affinity calculation unit 54 calculates the affinity as low if the human is elderly or female.

[0124] The concession operation planning unit 55 plans the evasive action of the autonomous mobile robot 51 and the anticipated evasive action of the human based on the affinity calculated by the affinity calculation unit 54, the operational goal generated by the robot operational goal generation unit 10, and the human operation predicted by the human operation prediction unit 11. If the affinity is high, there is no concern that the human's sense of ease will be compromised, even if the robot is relatively close to the human or approaches at high speed. Therefore, the concession operation planning unit 55 can create an efficient operational plan that allows for contact.If the affinity is low, the concession operations planning unit 55 creates an operations plan in which the speed is reduced when the robot is as far away from the human as possible or approaches the human without causing the human a feeling of discomfort.

[0125] The alarm operations planning unit 56 plans the alarm operation based on the affinity calculated by the affinity calculation unit 54 and the expected evasive operation planned by the concession operations planning unit 55. If the affinity is high, the alarm operations planning unit 56 plans a strong assertion alarm operation including contact. If the affinity is low, since an alarm via contact can disrupt a person's sense of ease or cause discomfort, the operations planning unit plans a weak assertion alarm operation using non-contact communication such as sound or light.For example, it is preferred that when affinity is high, a biological alarm such as speech or voice, contact or gesture—that is, an alarm familiar to humans—is selected, and when affinity is low, a mechanical alarm such as a warning tone or light is selected. In this case, a comfortable alarm can be implemented. With a comfortable alarm, an alarm can be implemented more efficiently without causing discomfort to the person. The description for the same case will not be repeated here.

[0126] Next, a control procedure of the control device 52 of the fifth embodiment will be explained.

[0127] As in Fig. As shown in Figure 10, the robot operation target generation unit 10 in the control unit 52 first performs an operation target generation process to generate the operation target (S71). Next, the human operation prediction unit 11 determines whether or not a person is detected in the environment, based on the position information of the obstacle from the radar sensor 6 and the image information from the camera 7 (S72). If it is determined that no person is detected, the human operation prediction unit 11 proceeds to step S75.

[0128] If it is determined that a human is being detected, the human operation prediction unit 11 performs operation prediction processing to predict the operation of the detected human (S73). Subsequently, the operation interference determination unit 15 determines whether or not the surrounding person interferes with the operation of the autonomous mobile robot 51, based on the operation target of the robot operation target generation unit 10 and the human operation predicted by the human operation prediction unit 11 (S74). If it is determined that no operation interference is present, the operation interference determination unit 15 proceeds to step S75.

[0129] In step S75, the motion control unit 13 performs normal robot control processing to control the motion mechanism of the autonomous moving robot 51 along the operation target generated by the robot operation target generation unit 10.

[0130] In step S74, when the operation interference determination unit 15 determines that an operation interference is present, the affinity calculation unit 54 performs an affinity calculation processing to calculate the affinity of the human for the autonomous moving robot 51 on the basis of the image information from the camera 7 (S76).

[0131] Subsequently, the concession operation planning unit 55 performs concession operation planning processing based on the affinity calculated by the affinity calculation unit 54, the operation goal generated by the robot operation goal generation unit 10, and the human operation predicted by the human operation prediction unit 11 (S77). The concession operation planning unit 55 plans the evasive action of the autonomous mobile robot 51 and the expected evasive action anticipated by the human as the concession operation. Then, the alarm operation planning unit 56 performs alarm operation planning processing to plan the alarm operation for the human based on the affinity calculated by the affinity calculation unit 54 and the expected evasive action planned by the concession operation planning unit 55 (S78).

[0132] In step S79, the motion control unit 13 and the alarm control unit 14 perform interference-assisted robot control processing to implement the concession operation. In this interference-assisted robot control processing, the motion control unit 13 controls the movement mechanism based on the avoidance operation of the autonomous moving robot 51, which is planned by the concession operation planning unit 55. The alarm control unit 14 controls the autonomous moving robot 51 to execute the alarm operation, which is planned by the alarm operation planning unit 56. Subsequently, the control unit 52 determines whether the operation goal is achieved and the operation of the autonomous moving robot 51 ends (S80). If it is determined that the operation of the autonomous moving robot 51 does not end, the control unit 52 repeats the processing starting from step S71.When it is determined that the operation of the autonomous mobile robot 51 is ending, the control device 52 terminates the control.

[0133] According to the autonomous mobile robot 51 of the fifth embodiment, operations are planned based on the human's affinity for the autonomous mobile robot 51, whereby the concession operation or the alarm operation can be executed without causing the human a feeling of discomfort. The alarm operation is planned based on the human's affinity for the autonomous mobile robot 51, whereby it is possible to prevent the human from experiencing a feeling of discomfort. [Sixth embodiment]

[0134] As in Fig. As shown in Figure 11, an autonomous mobile robot 61 of a sixth embodiment differs from the autonomous mobile robot 51 of the fifth embodiment in that a priority relationship determination unit 64, which determines the priority relationship between the human and the autonomous mobile robot 61, is provided, and the concession operation and the alarm operation are planned on the basis of the priority relationship determined by the priority relationship determination unit 64.

[0135] In the autonomous mobile robot 61 of the sixth embodiment, the concession operation and the alarm operation are planned using the priority relationship between the human and the autonomous mobile robot 61 and the human's affinity for the autonomous mobile robot 61.

[0136] The priority relationship determination unit 64 of the control device 62 of the sixth embodiment determines the priority relationship between the human interfering with the operation of the autonomous moving robot 61 and the autonomous moving robot 61, based on the position information of the obstacle from the radar sensor 6 and the image information from the camera 7.

[0137] The concession operation planning unit 65 plans the evasive operation of the autonomous mobile robot 61 and the expected evasive operation of the human based on the affinity calculated by the affinity calculation unit 54, the operation goal generated by the robot operation goal generation unit 10, the human operation predicted by the human operation prediction unit 11, and the determination result of the priority relationship determination unit 64. Similarly, the alarm operation planning unit 26 plans the alarm operation based on the affinity calculated by the affinity calculation unit 54, the expected evasive operation planned by the concession operation planning unit 65, and the determination result of the priority relationship determination unit 64.

[0138] Next, a control procedure of the control device 62 of the sixth embodiment will be explained.

[0139] As in Fig. As shown in Figure 12, the robot operation target generation unit 10 in the control unit 62 first performs an operation target generation process to generate the operation target (S91). Next, the human operation prediction unit 11 determines whether or not a person is detected in the environment, based on the position information of the obstacle from the radar sensor 6 and the image information from the camera 7 (S92). If it is determined that no person is detected, the human operation prediction unit 11 proceeds to step S95.

[0140] If it is determined that a human is being detected, the human operation prediction unit 11 performs operation prediction processing to predict the operation of the detected human (S93). Subsequently, the operation interference determination unit 15 determines whether or not the surrounding person interferes with the operation of the autonomous mobile robot 61, based on the operation target of the robot operation target generation unit 10 and the human operation predicted by the human operation prediction unit 11 (S94). If it is determined that no operation interference is present, the operation interference determination unit 15 proceeds to step S95.

[0141] In step S95, the motion control unit 13 performs normal robot control processing to control the motion mechanism of the autonomous moving robot 61 along the operation target generated by the robot operation target generation unit 10.

[0142] In step S94, when the Operational Interference Determination Unit 15 determines that operational interference is present, the Priority Relationship Determination Unit 64 performs priority relationship determination processing to determine the priority relationship between the autonomous mobile robot 61 and the human (S96). Subsequently, the Affinity Calculation Unit 54 performs affinity calculation processing to calculate the affinity based on the image information from the camera 7 (S97).

[0143] Subsequently, the concession operation planning unit 65 performs concession operation planning processing of the determination result of the priority relationship determination unit 64, the affinity calculated by the affinity calculation unit 54, the operation goal generated by the robot operation goal generation unit 10, and the human operation predicted by the operation prediction unit 11 (S98). The concession operation planning unit 65 plans, as the concession operation, the avoidance operation of the autonomous mobile robot 61 and the expected avoidance operation anticipated by the human interfering with the operation of the autonomous mobile robot 61.Subsequently, the Alarm Operations Planning Unit 66 performs an Alarm Operations Planning process to plan the alarm operation for the human based on the determination result of the priority relationship determination unit 64, the affinity calculated by the affinity calculation unit 54, and the expected evasion operation planned by the concession operations planning unit 65 (S99).

[0144] In step S100, the motion control unit 13 and the alarm control unit 14 perform interference-assisted robot control processing to implement the concession operation. In this interference-assisted robot control processing, the motion control unit 13 controls the movement mechanism based on the avoidance operation of the autonomous moving robot 61, which is planned by the concession operation planning unit 65. The alarm control unit 14 controls the autonomous moving robot 61 to execute the alarm operation, which is planned by the alarm operation planning unit 66.

[0145] The control unit 62 then determines whether the operational goal is achieved and the operation of the autonomous moving robot 61 ends (S101). If it is determined that the operation of the autonomous moving robot 61 does not change, the control unit 62 repeats the processing starting from step S91. If it is determined that the operation of the autonomous moving robot 61 ends, the control unit 62 terminates the control process.

[0146] According to the autonomous mobile robot 61 of the sixth embodiment, the same effects can be achieved as for the autonomous mobile robots of the second and fifth embodiments. The concession operation and the alarm operation are planned based on the priority relationship between the human and the autonomous mobile robot 61 and the human's affinity for the autonomous mobile robot 61, whereby an appropriate operational plan can be created according to the circumstances without causing the human a feeling of discomfort. [Seventh embodiment]

[0147] In a seventh embodiment, the invention is explained using a manipulator robot 71 as the movable body. The manipulator robot 71 performs a cooperative task in accordance with the operator and is, for example, a robot that supports the operator's task in a cell manufacturing system. The manipulator robot 71 moves the operator's hand to perform task support, such as holding or moving a workpiece or transporting components or tools, and performs task support in accordance with the operator's movements, thereby reducing obstacles for the operator or improving task efficiency. The manipulator robot 71 can be a mobile manipulator that moves on the floor or a fixed manipulator whose base is fixed to the floor.

[0148] In task support, a problem arises that even if a robot's support operation is synchronized with the operator's movement, the support the robot is supposed to perform may not be executed appropriately depending on the operator's task method (task procedure, task speed, way of holding an object, etc.). In particular, if the operator is replaced or the same task is performed in parallel across multiple manufacturing cells, the ability to adapt a setup or control program for each operator is limited. Consequently, after the robot has complied with a procedure requested by the operator, a new task procedure is scheduled, and an alarm is triggered to guide the operator.This alarm allows the operator to understand the type of movement expected by the robot, ensuring conformity with a task procedure anticipated by the robot. Consequently, a cooperative task is performed in a manner where both the human and the robot are aligned. When the robot's support is valid, tasks, quality, and efficiency are improved when the human cooperates with the robot to some extent, and appropriate task support is more readily achieved than when the human moves freely and the robot cannot provide assistance. Therefore, product quality or productivity can be improved.

[0149] In a cell manufacturing system suitable for producing a wide variety of products in small batches, one advantage is the ease with which product elements and specifications can be modified. However, if changes are frequent or made immediately after a change is made to a task, errors in the component type being assembled or a missing task procedure are likely. To address this problem, the robot is given high priority and guides the operator to the correct manufacturing component or task procedure, thereby reducing the occurrence of errors or missing tasks.Accordingly, in the case of the manipulator robot 71, as well as in the task support, if the operator performs an operation deviating from a plan, the alarm operation is executed in such a way that the operator follows a correct task method, thereby improving task quality and efficiency.

[0150] As in Fig. As shown in Figure 13, the manipulator robot 71 includes a control unit that performs overall control. The control unit 72 is connected to a position sensor 73, an attached camera 74 for operator use, and a contact sensor 75.

[0151] The position sensor 73 detects the position (i.e., the position of the joint and the hand or gripper) of the manipulator robot 71. The position sensor 73 transmits the detected position information to the control unit 72. The fixed camera 74 for operator use is a camera that records the state of an operator or of a task object to be supported by the manipulator robot 71. The fixed camera 74 for operator use transmits image information of the operator or the task object to the control unit 72. The contact sensor 75 detects a contact state of the manipulator robot 71. The contact sensor 75 transmits the information of the detected contact state to the control unit 72.

[0152] The control unit 72 includes a task support target generation unit 76, an operator operation prediction unit 77, an operation planning unit 78, a motion control unit 79 and an alarm control unit 80.

[0153] The task support target generation unit 76 generates a task support target for the manipulator robot 71. The task support target generation unit 76 generates the task support target, such as a motion target, based on the position information from the position sensor 73.

[0154] The operator operation prediction unit 77 detects and predicts the operator's operation based on the image information from the fixed camera 74 for operator use and the contact state information from the contact sensor 75.

[0155] The operations planning unit 78 has an operator operations determination unit 81, a priority relationship determination unit 82, a cooperation task operations planning unit 83 and an alarm operations planning unit 84.

[0156] The operator operation determination unit 81 determines whether or not the operator's operation has appropriate task content by comparing the operation predicted by the operator operation prediction unit 77 with a database containing a standard task. If the operator's operation is one that does not have appropriate task content, this refers, for example, to a task procedure that violates a predetermined standard specification, an operation that misuses a tool or component to be assembled, or the like.If it is determined that the operator's operation does not have adequate task content, the operator operation determination unit 81 determines whether or not the operator's operation can provide adequate support for the operator's operation, based on the task support target generated by the task support target generation unit 76 and the operation predicted by the operator operation prediction unit 77.If adequate support cannot be provided, this refers, for example, to a case where the position or orientation of the operator or the task object is turned away from or distant from the manipulator robot 71, or is located in a direction with respect to which the manipulator robot 71 has structural difficulties in providing support, an operation in which the operator's operational speed is excessively high or low, a case in which the operator's access position to the task object is inappropriate, or the like.

[0157] If the operator operation determination unit 81 determines that the operator's operation lacks adequate task content, the priority relationship determination unit 82 determines that the priority relationship between the operator and the manipulator robot 71 assigns its priority to the manipulator robot 71. If the operator operation determination unit 81 determines that the operator's operation has adequate task content and adequate support can be provided for the operator's operation, the priority relationship determination unit 82 determines the priority relationship based on the task support goal generated by the task support goal generation unit 76 and the operation predicted by the operator operation prediction unit 77. In this case, it is preferable to determine the priority according to the degree of contribution of a task support content.For example, if support content can significantly reduce the operator's workload or significantly improve task accuracy, manipulator robot 71 is assigned high priority. Alternatively, it is preferred to determine the priority according to the operator's skill level. If the operator's skill level is low, priority is assigned to manipulator robot 71, enabling it to provide task support that guides the human operator and allows for the use of island production, while preventing deterioration in quality or deviations, even without causing a significant decrease in production efficiency for a less skilled operator.If the operator's competence level is high, priority is assigned to the operator, with most task methods conforming to a method requested by the operator. If the operator does not want task support from manipulator robot 71, support is not forcibly provided, and it is possible to provide task support only if the operator requests it.

[0158] If the Operator Operation Determination Unit 81 determines that the operator's operation lacks adequate task content, the Cooperative Task Operation Planning Unit 83 schedules an operation that follows a standard task as an expected operation anticipated by the operator. If it is determined that adequate support for the operator's operation is not provided, the Cooperative Task Operation Planning Unit 83 schedules a cooperative task operation based on the determination result of the Priority Relationship Determination Unit 82, the task support goal generated by the Task Support Goal Generation Unit 76, and the operation predicted by the Operator Operation Prediction Unit 77.The support operation includes an evasive maneuver if the operator and the manipulator robot 71 collide, and the anticipated operation includes an anticipated evasive maneuver expected by the operator. In the cooperation task operation planning unit 83, an operation plan is created such that contact between the operator and the manipulator 71 is permitted. The contact tolerance in the seventh embodiment can have a fixed value or can be changed depending on the task content or the operator's task state. That is, for example, while the operator is performing an accurate task such as assembling narrow components, which requires high precision, or performing a task that handles dangerous objects such as a blade, the contact tolerance is reduced, making it possible to prevent contact from interfering with the operator's task and endangering the operator.

[0159] The alarm operation planning unit 84 plans an alarm operation that guides the operator, performs an operation that follows a standard task, or performs an operation that can be supported by the task support of the manipulator robot 71, based on the determination result of the priority relationship determination unit 82 and the expected operation planned by the cooperation task operation planning unit 83. In this case, the alarm operation is, for example, a procedure in which force is applied to the operator by contact to guide direct movement, a procedure in which an excessively high or low operation speed is communicated by sound generation rhythm or light flashing rate, or the like.

[0160] The motion control unit 79 controls the position of the manipulator robot 71 based on the support operation planned by the cooperation task operation planning unit 83. The alarm control unit 80 executes the alarm operation directed at the operator using communication methods such as voice output, light output, or contact, based on the alarm operation planned by the alarm operation planning unit 84.

[0161] Next, a control procedure of the control device 72 of the seventh embodiment will be explained.

[0162] As in Fig. As shown in Figure 14, the task support target generation unit 76 in the control unit 72 first performs task support target generation processing to generate the task support target (S111). Next, the operator operation prediction unit 77 predicts the operator's operation based on the image information from the fixed camera 74 for operator use and the contact state information from the contact sensor 75 (S112). Afterward, the operator operation determination unit 81 compares the operation predicted by the operator operation prediction unit 77 with the database that stores the standard task to determine whether or not the operator's operation has appropriate task content (S113).If it is determined that the operator's operation has appropriate task content, the operator operation determination unit 81 determines whether or not appropriate support can be provided for the operator's operation, based on the task support goal generated by the task support goal generation unit 76 and the operation predicted by the operator operation prediction unit 77 (S114). If it is determined that appropriate support can be provided for the operator's operation, the operator operation determination unit 81 proceeds to step S115.

[0163] In step S115, the motion control unit 79 performs normal robot control processing to control the position of the manipulator robot 71 along the task support target generated by the task support target generation unit 76.

[0164] In step S113, when the operator operation determination unit 81 determines that the operator's operation does not have adequate task content, or in step S114, when it is determined that measured support for the operator's operation cannot be provided, the priority relationship determination unit 82 performs priority relationship determination processing to determine the priority relationship between the human and the manipulator robot 71 (S116).

[0165] Subsequently, the Cooperative Task Operation Planning Unit 83 performs a Cooperative Task Operation Planning process to plan the cooperative task operation based on the determination result of the priority relationship determination unit 82, the task support goal generated by the task support goal generation unit 76, and the operation predicted by the operator operation prediction unit 77 (S117). The Cooperative Task Operation Planning Unit 83 plans the support operation as the cooperative task operation and the expected operation anticipated by the operator.Subsequently, the Alarm Operation Planning Unit 84 performs an alarm operation planning processing to plan the alarm operation that guides the operator to perform an operation following the standard task or to perform an operation to be supported by the task support of the manipulator robot 71, based on the determination result of the priority relationship determination unit 82 and the expected operation planned by the cooperation task operation planning unit 83 (S118).

[0166] In step S119, the motion control unit 79 and the alarm control unit 80 perform alarm-accompanied robot control processing to implement the cooperative task operation. In this alarm-accompanied robot control processing, the motion control unit 79 performs the position control of the manipulator robot 71 based on the support operation planned by the cooperative task operation planning unit 83. The alarm control unit 80 controls the manipulator robot 71 or an external audio device based on the alarm operation planned by the alarm operation planning unit 84 to execute the alarm operation.

[0167] The control unit 72 then determines whether the support goal has been reached and the operation of the manipulator robot 71 ends (S120). If it is determined that the operation of the manipulator robot 71 does not end, the control unit 72 repeats the processing starting from step 111. If it is determined that the operation of the manipulator robot 71 ends, the control unit 72 terminates the control.

[0168] According to the seventh embodiment of the manipulator robot 71, if the operator performs an operation that deviates from the plan, the alarm operation is executed in such a way that a correct task procedure is carried out, thereby reducing product defects or the omission of a task. Even if an individual operator uses an easy-to-operate task procedure, the operator's operation is guided in such a way that the manipulator robot 71 can provide task support, enabling efficient cooperative tasking and improving efficiency in the island manufacturing system. [Eighth embodiment]

[0169] In an eighth embodiment, the explanation is provided using an automobile 91 as the moving body of the invention. The automobile 91 is a vehicle capable of performing an automatic driving function to drive autonomously to a predetermined destination. The automobile 91 differs significantly from the autonomous moving robot 1 in that contact between a pedestrian and the automobile is not permitted. For this reason, the automobile 91 employs a concession operation to avoid contact with the pedestrian. The automobile 91 employs an alarm operation directed at the pedestrian using contactless communication via speech or light.

[0170] The priority relationship between the car 91 and the pedestrian generally gives priority to the pedestrian. However, this priority relationship is modified depending on road conditions, the pedestrian's condition, and surrounding circumstances, allowing the autonomously driven car 91 to pass through an environment where the pedestrian is present. Specifically, if a pedestrian moves away from one side of a road, if several pedestrians walk side by side, if a pedestrian walks on a roadway with a sidewalk, or similar situations, and if the pedestrian is obstructing the safe passage of the car 91, the pedestrian's priority is reduced and an alarm is triggered to expedite the pedestrian's movement, thus ensuring safe passage.If a pedestrian is an elderly person or a child, or if a pedestrian is pushing a wheelchair or stroller, or carrying a heavy load, it is determined that the pedestrian has high priority and no alarm, or only a discreet alarm, is sounded, and the vehicle 91 proceeds after the pedestrian moves to a safe position. If the vehicle 91 is an ambulance or similar vehicle, it is determined that the vehicle 91 has high priority and a evasive direction is communicated to the pedestrian by a positive alarm, allowing the pedestrian to proceed safely while making way for the vehicle 91, even in a busy location.

[0171] As in Fig. As shown in Figure 15, the automobile 91 includes a control unit 92 that performs the overall control. The control unit 92 is connected to a communication unit 93, a position detection unit 94, a map information storage unit, a radar sensor 96, and a camera 97.

[0172] The communication unit 93 receives a course plan for automatic driving via wireless communication with an external management center. The communication unit 93 transmits the received course plan for automatic driving to the control unit 92. The position detection unit 94 receives GPS signals to determine the current location of the vehicle 91. The position detection unit 94 determines the current location of the vehicle 91 based on map information stored in the map information storage unit 95. The position detection unit 94 transmits the information about the current location of the vehicle 91 and map information for the area surrounding the current location to the control unit 92.

[0173] The radar sensor 96 detects an obstacle in the vicinity of the vehicle 91 by means of reflected waves of emitted electrical waves. The radar sensor 96 transmits position information of the detected obstacle to the control unit 92. The camera 97 records the surroundings of the vehicle 91. The camera 97 transmits image information of the surroundings of the vehicle 91 to the control unit 92.

[0174] The control unit 92 comprises an automatic driving course generation unit 98, a pedestrian course prediction unit 99, an operations planning unit 100, a driving control unit 101, and an alarm control unit 102. The automatic driving course generation unit 98 generates an automatic driving course based on the automatic driving course plan from the communication unit 93, the current location information of the vehicle 91, and the map information for the vicinity of the current location from the position detection unit 94. The pedestrian course prediction unit 99 determines whether or not a pedestrian is detected in the vicinity based on the obstacle's position information from the radar sensor 96 and the image information from the camera 97.If it is determined that a pedestrian is detected, the pedestrian course prediction unit 99 predicts the course of the detected pedestrian based on the position information of the obstacle from the radar sensor 96 and the image information from the camera 97.

[0175] The operations planning unit 100 includes a safety area determination unit 103, a priority relationship determination unit 104, an affinity calculation unit 105, a concession course planning unit 106, and an alarm operations planning unit 107.

[0176] The safety area determination unit 103 determines whether or not the safety area, such as ensuring that the pedestrian does not interfere with the course of the automobile 91, can be ensured, based on the course for automatic driving generated by the automatic driving course generation unit 98 and the course of the pedestrian predicted by the pedestrian course prediction unit 99.

[0177] If the safety area determination unit 103 determines that the safety area cannot be ensured, the priority relationship determination unit 104 determines the priority relationship between the pedestrian and the vehicle 91. The priority relationship determination unit 104 determines the priority relationship between the pedestrian and the vehicle 91 based on the obstacle's position information from the radar sensor 96 and the image information from the camera 97. If the pedestrian interferes with the safe passage of the vehicle 91, the priority relationship determination unit 104 determines that the pedestrian has a low priority. If the vehicle is in an emergency situation, the priority relationship determination unit 104 determines that the vehicle 91 has a high priority.

[0178] Affinity Calculation Unit 105 calculates the pedestrian's affinity for the automobile 91. Affinity Calculation Unit 105 calculates the affinity based on image information from camera 97. Affinity Calculation Unit 105 calculates the affinity based on the type of automobile 91, its apparent size, or similar factors. Specifically, if the automobile 91 is a heavy truck or bus, Affinity Calculation Unit 105 calculates the affinity as low. If the automobile 91 is a compact car, Affinity Calculation Unit 105 calculates the affinity as high.

[0179] The concession course planning unit 106 plans a concession course based on the determination result of the priority relationship determination unit 104, the affinity calculated by the affinity calculation unit 105, the course for an automated driving system generated by the course generation unit 98, and the pedestrian's course predicted by the pedestrian course prediction unit 99. The concession course planning unit 106 plans as the concession course an alternative course for the automobile 91 and an expected alternative course anticipated by the pedestrian.

[0180] The alarm operation planning unit 107 plans an alarm operation that guides the pedestrian to the alternative course, based on the expected alternative course planned by the concession course planning unit 106, the determination result of the priority relationship determination unit 104, and the affinity calculated by the affinity calculation unit 105.

[0181] The driving control unit 101 automatically drives the car 91 based on the alternative course planned by the concession course planning unit 106. The alarm control unit 102 carries out the alarm operation for or on the pedestrian by voice or light based on the alarm operation planned by the alarm operation planning unit 107.

[0182] Next, a control procedure of the control device 92 of the eighth embodiment will be explained.

[0183] As in Fig. As shown in Figure 16, the first step in the control unit 92 is for the generation unit 98 to perform course generation processing to generate the course for automatic driving (S131). Next, the pedestrian course prediction unit 99 determines whether or not a pedestrian is detected in the vicinity, based on the position information of the obstacle from the radar sensor 96 and the image information from the camera 97 (S132). If it is determined that no pedestrian is detected, the pedestrian course prediction unit 99 proceeds to step S335. If it is determined that a pedestrian is detected, the pedestrian course prediction unit 99 performs pedestrian course prediction processing to predict the pedestrian's path (S133).

[0184] The safety zone determination unit 103 then determines whether or not the safety zone can be ensured in such a way that the pedestrian does not interfere with the course of the automobile 91, based on the course for automatic driving generated by the course generation unit 98 and the course of the pedestrian predicted by the pedestrian course prediction unit 99 (S134). If it is determined that the safety zone cannot be ensured, the safety zone determination unit 103 proceeds to step S135.

[0185] In step S135, the driving control unit 101 performs normal vehicle control processing for the automatic driving of the automobile 91 along the course for automatic driving, which is generated by the generating unit 98 for a course for automatic driving.

[0186] In step S134, if the safety area determination unit 103 determines that the safety area cannot be ensured due to course interference, the priority relationship determination unit 104 determines the priority relationship between the pedestrian and the car 91 (S136). Then, the affinity calculation unit 105 calculates the affinity of the pedestrian for the car 91. The affinity calculation unit 105 calculates the affinity based on the image information from the camera 97 (S137).

[0187] Subsequently, the concession course planning unit 106 performs a concession course planning processing to plan the concession course based on the determination result of the priority relationship determination unit 104, the affinity calculated by the affinity calculation unit 105, the course for automatic procedure generated by the generation unit 98 for an automatic driving course, and the course of the pedestrian predicted by the pedestrian course prediction unit 99 (S138).

[0188] The Alarm Operations Planning Unit 107 performs an alarm operations planning processing to plan the alarm operation that will guide the pedestrian to the expected diversion route, based on the expected diversion course planned by the concession course planning unit 106, the determination result of the priority relationship determination unit 104, and the affinity calculated by the affinity calculation unit 105 (S139).

[0189] In step S140, the driving control unit 101 and the alarm control unit 102 perform interference-assisted vehicle control processing to implement concessions. During this interference-assisted vehicle control processing, the driving control unit 101 executes automatic driving based on the evasive course of the car 91, which is planned by the concession course planning unit 106. The alarm control unit 102 controls a loudspeaker or a light of the car 91 based on the alarm operation planned by the alarm operation planning unit 107, in order to execute the alarm operation by voice or light.

[0190] The control unit 92 then determines whether or not the automatic procedure for the automated driving course ends and the vehicle arrives at the destination (S141). If it is determined that the vehicle 91 does not arrive at the destination, the control unit 92 repeats the processing starting from step S131. If it is determined that the vehicle 91 arrives at the destination, the control unit 92 ends the control process.

[0191] According to the eighth embodiment of Automobile 91, concessions are made to the pedestrian, allowing the pedestrian to walk safely even in a busy environment while making way for the automobile. The route is planned by considering the priority relationship or affinity, whereby an appropriate course can be planned according to the circumstances of the pedestrian and Automobile 91, while causing as little discomfort as possible to the pedestrian. [Ninth embodiment]

[0192] In a ninth embodiment, a description is provided using a personal mobility device 111 as the movable body of the invention. The personal mobility device 111 is a single passenger robot onto which a person steps and performs a driving operation.

[0193] It is assumed that the personal mobility device operates in a location such as a pedestrian walkway or sidewalk, or within a building, where pedestrians are the primary users. However, in a busy environment, a problem arises: allowing the personal mobility device to operate without causing a disturbance to the surroundings is not straightforward, and a certain level of driving skill is required. Therefore, Personal Mobility Device 111 incorporates a concessionary operation, ensuring safe movement even by a person without driving experience in a busy environment.

[0194] As in Fig. As shown in Figure 17, the personal mobility device 111 includes a control unit 112 that performs the overall control. The control unit 112 is connected to a driving operation input unit 113, a radar sensor 114, and a camera 115.

[0195] The driving operation input unit 113 is a device used when a driver inputs a driving operation into the personal mobility device 111. The driving operation input unit 113 is configured, for example, as a control lever, a handle, a switch, or a pedal. The driving operation input unit 113 outputs a driving operation from the driver to the control unit 112. A case in which the driver does not move onto the personal mobility device 111 is compensated for, and in this case, remote operation is carried out by communication using a remote control device or the like. In the ninth embodiment, the personal mobility device 111 moves using a driving operation input as an operation target.

[0196] The radar sensor 114 detects an obstacle in the vicinity of the personal mobility device 111 by means of reflected waves of emitted electrical waves. The radar sensor 114 transmits position information of the detected obstacle to the control unit 112. The camera 114 records the area around the personal mobility device 111. The camera 115 transmits image information of the area to the control unit 112.

[0197] The control unit 112 includes a pedestrian course prediction unit 116, an operations planning unit 117, a motion control unit 118 and an alarm control unit 119.

[0198] The pedestrian course prediction unit 116 determines whether or not a surrounding pedestrian is detected, based on the obstacle's position information from the radar sensor 114 and the image information from the camera 115. If it is determined that the surrounding pedestrian is detected, the pedestrian course prediction unit 116 predicts the pedestrian's course based on the obstacle's position information from the radar sensor 114 and the image information from the camera 115.

[0199] The operations planning unit 117 comprises a course interference determination unit 120, a priority relationship determination unit 121, an affinity calculation unit 122, a concession course planning unit 123, and an alarm operations planning unit 124. The course interference determination unit 120 determines whether or not the pedestrian interferes with (crosses or approaches) the course of the personal mobility device 111 based on the driving operation output from the driving operation input unit 113 and the pedestrian's course predicted by the pedestrian course prediction unit 116.

[0200] If the course interference determination unit 120 determines that the pedestrian is interfering with the course of the personal mobility device 111, the priority relationship determination unit 120 determines the priority relationship between the pedestrian and the personal mobility device 111. The priority relationship determination unit 121 determines the priority relationship based on the obstacle's position information from the radar sensor 114 and the image information from the camera 115. The personal mobility device 111 may be equipped with a selector switch or a setting knob that inputs the driver's urgency level to the control unit 112. In this case, the priority relationship determination unit 121 determines the priority relationship based on the input urgency level.The affinity calculation unit 122 calculates the pedestrian's affinity for the personal mobility device 111 based on the image information from the camera 115.

[0201] The concession course planning unit 123 plans a concession course based on the determination result of the priority relationship determination unit 121, the affinity calculated by the affinity calculation unit 122, the driving operation output by the driving operation input unit 113, and the pedestrian's course predicted by the pedestrian course prediction unit 116. The concession course planning unit 123 plans as the concession course an alternative course of the personal mobility device 111 and an expected alternative course anticipated by the pedestrian.

[0202] The alarm operation planning unit 124 plans an alarm operation that guides the pedestrian to the expected alternative course, based on the expected alternative course planned by the concession course planning unit 123, the determination result of the priority relationship determination unit 122, and the affinity calculated by the affinity calculation unit 122.

[0203] The motion control unit 118 controls the movement mechanism of the personal mobility device 111 based on the evasive course planned by the concession course planning unit 123. The alarm control unit 119 executes the alarm operation directed at the pedestrian by voice, light, or contact based on the alarm operation planned by the alarm operation planning unit 124. Next, a control procedure of the control device 112 of the ninth embodiment is described.

[0204] As in Fig. As shown in Figure 18, the first step in the control unit 112 is to input the driver's driving operation from the driving operation input unit 113 (S151). Next, the pedestrian course prediction unit 116 determines whether or not a surrounding pedestrian is detected, based on the obstacle's position information from the radar sensor 114 and the image information from the camera 115 (S152). If it is determined that no surrounding pedestrian is detected, the pedestrian course prediction unit 116 proceeds to step S155. If it is determined that a surrounding pedestrian is detected, the pedestrian course prediction unit 116 performs pedestrian course prediction processing to predict the pedestrian's path (S153).

[0205] The course interference determination unit 120 then determines whether or not the pedestrian interferes with the course of the personal mobility device 111, based on the driving operation output by the driving operation input unit 113 and the pedestrian's course predicted by the pedestrian course prediction unit 116 (S154). If it is determined that there is no interference with the pedestrian, the course interference determination unit 120 proceeds to step S155.

[0206] In step S155, the motion control unit 118 performs normal mobility control processing to control the movement mechanism of the personal mobility device 111 according to the driving operation output by the driving operation input unit 113.

[0207] In step S154, when the course interference determination unit 120 determines that interference with the pedestrian exists, the priority relationship determination unit 121 determines the priority relationship between the pedestrian and the personal mobility device 111 (S156). Then, the affinity calculation unit 122 calculates the pedestrian's affinity for the personal mobility device 111. The affinity calculation unit 122 calculates the affinity based on the image information from the camera 115 (S157).

[0208] Subsequently, the concession course planning unit 123 performs a concession course planning processing to plan the concession course based on the determination result of the priority relationship determination unit 121, the affinity calculated by the affinity calculation unit 122, the driving operation output by the driving operation input unit 113, and the pedestrian course predicted by the pedestrian course prediction unit 116 (S158).

[0209] The alarm operation planning unit 124 performs alarm operation planning processing to plan the alarm operation that guides the pedestrian to the expected alternative course, based on the expected alternative course planned by the concession course planning unit 123, the determination result of the priority relationship determination unit 121, and the affinity calculated by the affinity calculation unit 122 (S159).

[0210] In step S160, the motion control unit 118 and the alarm control units 119 perform interference-assisted mobility control processing to implement concessions. In this interference-assisted mobility control processing, the motion control unit 118 controls the movement mechanism based on the evasive course of the personal mobility device 111, which is planned by the concession course planning unit 123. The alarm control unit 119 executes the alarm operation using communication via voice output, light output, or contact, based on the alarm operation planned by the alarm operation planning unit 124.

[0211] The control unit 112 then determines whether or not the driving of the personal mobility device 111 ends (S161). If it is determined that the driving does not end, the control unit 112 repeats the processing starting from step S151. If it is determined that the driving ends, the control unit 112 terminates the control process.

[0212] According to the personal mobility device 111 of the ninth embodiment, concessions are made to the pedestrian, allowing the pedestrian to walk safely even in a busy traffic environment while making way for the automobile. The route is planned by considering the priority relationship or affinity, whereby an appropriate course can be planned according to the circumstances of the pedestrian and the personal mobility device 111, while causing as little discomfort as possible to the pedestrian. [Tenth embodiment]

[0213] As in Fig. As shown in Figure 19, an autonomous mobile robot 131 of a tenth embodiment differs from the autonomous mobile robot of the first embodiment mainly in that it provides an operation outcome prediction unit 134, which predicts the operation outcomes of several operation candidates, and an operation candidate selection unit 135, which selects an operation candidate from several operation candidates for which it is predicted that a predetermined boundary condition will be met.

[0214] A control unit 132 of the autonomous mobile robot 131 of the tenth embodiment includes an operation planning unit 133, which comprises the operation outcome prediction unit 134 and the operation candidate selection unit 135. The operation outcome prediction unit 134 predicts the operation outcomes when several planned operation candidates of the autonomous mobile robot 131 are executed in practice. The operation candidates of the autonomous mobile robot 131 are planned based on the evasive operation of the autonomous mobile robot 131, which is planned by the concession operation planning unit 125, and the expected evasive operation anticipated by the human operator. The operation candidates include the evasive operation and the alarm operation of the autonomous mobile robot 131.

[0215] The Operation Outcome Predictor Unit 134 predicts the operation outcomes using a human motion model (including a model of a motion path for the alarm operation of the robot 131). The Operation Outcome Predictor Unit 134 establishes the position and velocity as initial conditions for the robot 131 and the human as they cross paths. The Operation Outcome Predictor Unit 134 then predicts the operation outcomes by considering the acceleration of the robot 131 and the acceleration of the human 131 due to a contact force of the alarm operation to be performed by the robot 131. The Operation Outcome Predictor Unit 134 repeats the prediction for multiple operation candidates that differ in the intensity (contact force) of the alarm operation of the autonomous moving robot 131 or that differ in the parameters of the avoidance acceleration.

[0216] The operation candidate selection unit 135 determines whether or not the operation candidate for whom the operation outcome is predicted to satisfy the predetermined boundary condition can be selected, based on the prediction of the operation outcome by the operation outcome prediction unit 134. Examples of predetermined boundary conditions include the distance between the human and the robot 131, the intensity of an alarm (the intensity of an alarm by light or speech, or the intensity of an alarm by contact), and the like.

[0217] If it is determined that the operation candidate can be selected, the operation candidate selection unit 135 computes an operation candidate with the maximum motion efficiency from among those predicted to satisfy the predetermined boundary condition. Specifically, the operation candidate selection unit 135 evaluates the motion efficiency of the selectable operation candidate based on the prediction of the operation outcomes by the operation outcome prediction unit. The motion efficiency can be evaluated by focusing on the change in velocity before and after concessions. Since motion is maintained before concessions, and the change in velocity before and after concessions is small, the operation candidate selection unit 135 evaluates the motion efficiency as high.

[0218] The surgical candidate selection unit 135 evaluates the motion efficiency of the human and the motion efficiency of the robot 131. The surgical candidate selection unit 135 selects a surgical candidate whose sum of the motion efficiency of the human and the motion efficiency of the robot 131 is maximized.

[0219] Next, a control procedure of the control device 132 of the tenth embodiment will be explained.

[0220] As in Fig.As shown in Figure 20, the robot operation target generation unit 10 in the control unit 132 first performs an operation target generation process to generate the operation target (S171). Next, the human operation prediction unit 11 determines whether or not a human is detected in the environment, based on the position information of the obstacle from the radar sensor 6 and the image information of the robot's environment from the camera 7 (S172). If it is determined that no human is detected, the human operation prediction unit 11 proceeds to step S175.

[0221] If it is determined that a human is being detected, the human operation prediction unit 11 performs predictive processing to predict the operation of the detected human (S173). Subsequently, the operation interference determination unit 15 determines whether or not the surrounding person interferes with the operation of the autonomous mobile robot 131, based on the operation target from the robot operation target generation unit 10 and the human operation predicted by the human operation prediction unit 11 (S174). If it is determined that no operation interference is present, the operation interference determination unit 15 proceeds to step S175.

[0222] In step S175, the motion control unit 13 performs normal robot control processing to control the motion mechanism of the autonomous moving robot 131 along the operation target generated by the robot operation target generation unit 10.

[0223] In step S174, when the Operational Interference Determination Unit 15 determines that operational interference exists, the Concessional Operation Planning Unit 25 performs concessional operation planning processing based on the determination result of the priority relationship determination unit 24, the operational goal generated by the robotic operation goal generation unit 10, and the human operation predicted by the human operation prediction unit 11 (S176). The Concessional Operation Planning Unit 25 plans the concessional operation, the avoidance operation of the autonomous mobile robot 131, and the expected avoidance operation anticipated by the human.

[0224] Subsequently, the Alarm Operations Planning Unit 26 performs an Alarm Operations Planning Process to plan the alarm operation directed at the human on the basis of the determination result of the priority relationship determination unit 24 and the expected evasion operation planned by the concession operations planning unit 25 (S177).

[0225] The operational candidate is a combination of the contingency operation and the alert operation. The concessional operations planning unit 25 and the alert operations planning unit 26 plan multiple operational candidates from several contingency operations and alert operations.

[0226] Subsequently, the Operation Outcome Predictor Unit 134 predicts the operational outcomes when several planned operational candidates of the autonomous mobile robot 131 are performed in practice (S178). The Operation Outcome Predictor Unit 134 predicts the operational outcomes using the human motion model (including the model of the movement path in response to the alarm operation of the robot 131).

[0227] Next, the operation candidate selection unit 135 determines whether or not the operation candidate for which the operation outcome is predicted to satisfy the predetermined boundary condition can be selected based on the prediction of the operation outcomes by the operation outcome prediction unit 134 (S179).

[0228] If there is no operation candidate for which the predetermined boundary condition is predicted to be satisfied, the operation candidate selection unit 135 determines that the operation candidate is not selectable. In this case, processing continues to step S176 and a new operation candidate is generated. If an operation candidate exists for which the predicted operation result is predicted to satisfy the predetermined boundary condition, the operation candidate selection unit 135 determines that the operation candidate is selectable.

[0229] If it is determined that the surgical candidate is selectable, the surgical candidate selection unit 135 selects a surgical candidate with the maximum motion efficiency and the surgical candidate for which the predetermined boundary condition is predicted to be satisfied (S180). The surgical candidate selection unit 135 evaluates the motion efficiencies of several surgical candidates based on the prediction of the surgical outcomes by the surgical outcome prediction unit 134. The surgical candidate selection unit 134 evaluates the motion efficiency of the human and the motion efficiency of the robot 131 and selects the surgical outcome in which the sum of the motion efficiency of the human and the motion efficiency of the robot 131 is maximized.

[0230] Subsequently, in step S181, the motion control unit 13 and the alarm control unit 14 execute interference-assisted robot control processing to implement the concession operation based on the operation candidate selected by the operation candidate selection unit 135. In the interference-assisted robot control processing, the avoidance operation and the alarm operation are executed according to the operation candidate selected by the operation candidate selection unit 135.

[0231] The control unit 132 then determines whether the operational goal is achieved and the operation of the autonomous mobile robot 131 ends (S182). If it is determined that the operation of the autonomous mobile robot 131 does not end, the control unit 132 repeats the processing starting from step S131. If it is determined that the operation of the autonomous mobile robot 131 changes, the control unit 132 terminates the control process.

[0232] According to the autonomous mobile robot 131 of the tenth embodiment, the operation candidate for which it is predicted that the predetermined boundary condition, such as an approach to a human or an intensity of an alarm, will be met, is selected from several operation candidates, whereby it is possible to suppress an inappropriate operation, such as an excessively intense alarm.

[0233] In particular, if a person is expected to largely avoid evasive action, an intense alarm is necessary. However, from the standpoint of safety or well-being, it is not preferable to execute an excessively intense alarm operation. In the case of a contact-based alarm, if the alarm is excessively intense, the person may experience pain or their body position may become unstable. In the case of an audible or visual alarm, if the alarm is excessively intense, the stimulus may cause the person extreme discomfort. If there is a small probability that an evasive action can be expected from the person, the robot's movement efficiency will be reduced.

[0234] Accordingly, in the autonomous mobile robot 131 of the tenth embodiment, a search procedure is used which selects an operation candidate that fulfills a predetermined boundary condition and has high motion efficiency from several operation candidates, whereby a concession operation that has high motion efficiency and can be realized in practice can be performed.

[0235] The invention is not limited to the aforementioned embodiments.

[0236] For example, the movable body of the invention is not limited to an autonomous moving robot, a manipulator, an automobile, and a personal mobility device. The movable body is not necessarily limited to a moving body that moves on the ground and can be a moving body that moves in the air or underwater.

[0237] Only a section of the palm can be used as a region to trigger an alarm when the robot comes into contact with a person. This is because, in the case of contact via the palm section, the contact status can be detected with high accuracy by any sensor compared to contact via other regions. Contact via the palm section is therefore excellent because the impact force can be reduced by using a flexible covering for the palm section or a passive flexible mechanism for an arm section.

[0238] Although the preceding embodiments describe a case in which an operation is performed for a human as the other party, the invention can be suitably applied to parties other than humans. That is to say, the movable obstacle described in the claims includes an animal, such as a dog, a mobility device such as a car, or other robots in addition to a human. Industrial applicability

[0239] The invention can be used for a movable body such as a robot, a manipulator or a mobility device. Reference symbol list

[0240] 1, 21, 31, 41, 51, 61, 131: autonomous mobile robot (moving body), 2, 22, 32, 42, 52, 62, 72, 92, 112, 132, control unit, 6, 96, 114: radar sensor (means for detecting a moving body), 7, 97, 115: camera (means for detecting a moving body), 10: robot operation target generation unit (operation target generation means), 11: human operation prediction unit (operation prediction means), 12, 23, 34, 43, 53, 63, 78, 100, 117, 133: operation planning unit (operation planning means), 15: operation interference determination unit (proximity determination means), 16, 25, 36, 46, 55, 65: Concession Operations Planning Unit, 17, 26, 37, 47, 56, 66, 84, 100, 107, 124: Alarm Operations Planning Unit (Alarm Operations Planning Tool), 24, 44, 64, 82, 104, 121: Priority Relationship Determination Unit (Priority Relationship Determination Tool), 35, 45: Contact Tolerance Calculation Unit (Contact Tolerance Calculation Tool), 54, 105,122: Affinity Calculation Unit (Affinity Calculation Means), 71: Manipulator Robot (Moving Body), 74: Fixed Camera for Operator Use (Means of Detecting a Moving Obstacle), 76: Task Support Target Generation Unit (Operation Target Generation Means), 77: Operator Operation Predictor Unit (Operation Predictor Means), 81: Operator Operation Determinor Unit (Proximity Determinor Means), 83: Cooperative Task Operation Planning Unit (Operation Planning Means), 91: Automobile (Moving Body), 98: Auto-Driving Course Generation Unit (Operation Target Generation Means), 99, 116: Pedestrian Course Predictor Unit (Operation Predictor Means), 103: Safety Area Determinor Unit (Proximity Determinor Means), 106, 123: Concession Course Planning Unit (Operation Planning Means), 111: Personal Mobility Device (Moving Body), 120: Course interference determination unit (approximation method),134: Surgical Outcome Predictor Unit, 135: Surgical Candidate Selection Unit,

Images