JP7702616B2 - Autonomous traveling robot system, judgment adjustment method, and program - Google Patents

Description

This disclosure relates to an autonomous robot system, a judgment adjustment method, and a program.

Patent document 1 discloses a method for receiving three-dimensional image data from an imaging device, segmenting the received three-dimensional image data into objects, identifying a person corresponding to at least a portion of the filtered object, and causing a mobile robot to follow the identified person.

JP 2014-195868 A

However, with the method described in Patent Document 1, even if an erroneous identification result is derived when identifying objects and people, the identification result cannot be corrected or the identification method adjusted, so there is a high possibility that erroneous identification will be repeated.

The present disclosure provides an autonomous robot system, a judgment adjustment method, and a program that are less likely to repeatedly make the same erroneous judgment when determining whether an object is a person or an obstacle.

An autonomous mobile robot system according to one aspect of the present disclosure is an autonomous mobile robot system equipped with an autonomous mobile robot that autonomously travels on a predetermined floor, and includes a sensor unit that detects objects around the autonomous mobile robot and acquires sensing data related to the detected objects, a determination unit that determines whether the object is a person or an obstacle based on the sensing data, and an evaluation unit that evaluates the determination result by the determination unit to be an erroneous determination if the sensing data and the determination result are registered in an erroneous determination database as historical information of erroneous determination, and when the evaluation unit evaluates the determination result to be an erroneous determination, the determination unit at least either corrects the determination result or adjusts the method of determining the object.

A judgment adjustment method according to one aspect of the present disclosure is a judgment adjustment method executed by one or more computers, in which the one or more computers detect objects around an autonomous robot, acquire sensing data related to the detected objects, and determine whether the object is a person or an obstacle based on the acquired sensing data. If the sensing data and the judgment result are registered in an erroneous judgment database as historical information of an erroneous judgment, the judgment result is evaluated as an erroneous judgment. If the judgment result is evaluated as an erroneous judgment, the judgment result is corrected and/or the judgment method for the object is adjusted.

A program according to one aspect of the present disclosure is a program for causing one or more computers to execute the judgment adjustment method.

According to this disclosure, it is less likely that similar erroneous judgments will be made repeatedly when determining whether an object is a person or an obstacle.

FIG. 1 is a block diagram illustrating an example of a functional configuration of an autonomous mobile robot system according to an embodiment. FIG. 2 is a perspective view of the autonomous mobile robot according to the embodiment, seen obliquely from above. FIG. 3 is a bottom view showing the external appearance of the autonomous mobile robot according to the embodiment as viewed from the rear side. FIG. 4 is a diagram showing an evaluation of the correctness of the judgment result. FIG. 5 is a diagram showing an example of the operation of the autonomous mobile robot and a human when the human is erroneously determined to be an obstacle. FIG. 6 is a diagram showing an example of the operation of the autonomous mobile robot corresponding to the determination result. FIG. 7 is a diagram showing an example of data stored in the erroneous determination database. FIG. 8 is a diagram showing another example of data stored in the erroneous determination database. FIG. 9 is a diagram showing an example of data stored in the pattern database. FIG. 10 is a diagram showing an example of a display screen. FIG. 11A is a flowchart showing an example of the operation of the autonomous mobile robot system according to the embodiment. FIG. 11B is a flowchart showing an example of the operation of the autonomous mobile robot system according to the embodiment. FIG. 11C is a flowchart showing an example of the operation of the autonomous mobile robot system according to the embodiment. FIG. 12 is a diagram illustrating an example of a driving plan. FIG. 13 is a diagram illustrating the operation of the autonomous mobile robot when it determines that a detected object is an obstacle. FIG. 14 is a diagram showing a schematic operation of an autonomous mobile robot to avoid an obstacle. FIG. 15 is a diagram showing a schematic diagram of the behavior of a person when an autonomous mobile robot approaches the person. FIG. 16 is a diagram illustrating the operation of an autonomous mobile robot when an object is erroneously determined to be an obstacle. FIG. 17 is a diagram showing a schematic operation of the autonomous mobile robot when an object is erroneously determined to be an obstacle. FIG. 18 is a diagram illustrating the operation of the autonomous mobile robot when it determines that a detected object is a person. FIG. 19 is a diagram illustrating the operation of an autonomous mobile robot to avoid a person. FIG. 20 is a diagram showing a schematic operation of an autonomous mobile robot revisiting a position where it has avoided a person. FIG. 21 is a diagram showing an example of a cleaning result displayed on the display screen. FIG. 22 is a diagram showing an example of accepting an input operation for correcting a determination result.

The following describes the embodiments in detail with reference to the drawings. Note that the embodiments described below are all comprehensive or specific examples. The numerical values, shapes, materials, components, the arrangement and connection of the components, steps, and the order of steps shown in the following embodiments are merely examples and are not intended to limit the scope of the claims. Also, each figure is not necessarily an exact illustration. In each figure, substantially identical configurations are given the same reference numerals, and duplicate explanations may be omitted or simplified.

Note that each figure is a schematic diagram and is not necessarily a precise illustration. In addition, in each figure, the same reference numerals are used for substantially the same configurations, and duplicate explanations may be omitted or simplified.

In the following, terms indicating the shape of elements, such as approximately circular, approximately rectangular, and approximately polygonal, do not only indicate the strict meaning, but also include a range of substantial equivalence, for example, a difference of about a few percent.

(Embodiment)
An autonomous traveling robot system according to an embodiment will be described below. The autonomous traveling robot system according to an embodiment detects objects around an autonomous traveling robot autonomously traveling in a predetermined area, judges whether the detected object is a person or an obstacle, and at least one of correcting the judgment result and adjusting the judgment method for the object if the judgment result is an erroneous judgment. The autonomous traveling robot system evaluates whether the judgment result is erroneous based on an erroneous judgment database in which historical information on erroneous judgments is stored, and at least one of correcting the judgment result and adjusting the judgment method, so that the same erroneous judgment is less likely to be repeated. The configuration of the autonomous traveling robot system will be described below.

[1. Configuration]
1 is a block diagram showing an example of a functional configuration of an autonomous mobile robot system according to an embodiment. The autonomous mobile robot system 300 includes, for example, an autonomous mobile robot 100 and a terminal device 200. Each component will be described below.

[1-1. Autonomous Robot]
The autonomous mobile robot 100 is a robot that autonomously travels on a predetermined floor. In the embodiment, an example in which the autonomous mobile robot 100 is an autonomous mobile robot cleaner having a cleaning function will be described, but the present invention is not limited to this example. For example, the autonomous mobile robot 100 may have an item management, transportation, security, guidance, search, or disinfection function.

As shown in FIG. 1, the autonomous mobile robot 100 includes, for example, a communication unit 110, a control unit 120, a memory unit 130, a sensor unit 140, a notification unit 150, a traveling unit 160, and a cleaning unit 170.

The configuration of the autonomous mobile robot 100 will be described below with reference to Fig. 1, as well as Fig. 2 and Fig. 3. Fig. 2 is a perspective view of the autonomous mobile robot 100 in the embodiment, as seen diagonally from above. Fig. 3 is a bottom view showing the appearance of the autonomous mobile robot 100 in the embodiment, as seen from the rear side.

[Communications Department]
The communication unit 110 is a communication circuit for the autonomous mobile robot 100 to communicate with the terminal device 200. The communication unit 110 may include a communication circuit (communication module) for communicating via the wide area communication network 10 and a communication circuit (communication module) for communicating via a local communication network. The communication unit 110 is, for example, a wireless communication circuit for performing wireless communication. Note that there is no particular limitation on the communication standard for the communication performed by the communication unit 110.

[Control unit]
The control unit 120 performs various information processing for controlling the operation of the autonomous mobile robot 100. The control unit 120 is realized, for example, by a microcomputer or a microcontroller, but may be realized by a processor. The control unit 120 has, as functional components, for example, an acquisition unit 121, a self-position calculation unit 122, a map creation unit 123, a determination unit 124, an evaluation unit 125, an object position calculation unit 126, a travel plan creation unit 127, a travel control unit 128, and a cleaning control unit 129. The functions of these components are realized, for example, by the microcomputer, microcontroller, or processor constituting the control unit 120 executing a computer program stored in the storage unit 130.

[Acquisition part]
The acquisition unit 121 acquires sensing data acquired by the sensor unit 140 .

[Self-position calculation unit]
The self-position calculation unit 122 estimates a self-position indicating the current position of the main body 101 of the autonomous mobile robot 100 on a map. For example, the self-position calculation unit 122 calculates the current position of the autonomous mobile robot 100 on a map by, for example, SLAM, based on sensing data obtained from various sensors included in the sensor unit 140 described later and map information of a map created by the map creation unit 123. The various sensors will be described later.

[Map Creation Department]
The map creation unit 123 creates a map of a specific floor on which the autonomous mobile robot 100 travels. The map creation unit 123 may create a map of a specific floor using a map creation technique such as SLAM (Simultaneous Localization and Mapping), or may acquire map information input from an external device (not shown) via a network. The map information may be stored in advance in the storage unit 130, and in this case, the map creation unit 123 may read and acquire the map information from the storage unit 130.

The map creation unit 123 may also create a cleaning history map as a result of cleaning, for example. More specifically, the map creation unit 123 creates a cleaning history map by linking the travel path, the object determination result, the type and amount of garbage collected, and the like, onto a map showing a specific floor.

[Judgment section]
The determination unit 124 determines whether an object detected around the autonomous mobile robot 100 is a person or an obstacle based on the sensing data acquired by the sensor unit 140. More specifically, the determination unit extracts a feature amount included in the sensing data, and determines whether the object is a person or an obstacle based on the extracted feature amount. For example, when the sensor unit 140 is a LiDAR (Light Detection and Ranging), the feature amount is point cloud information, but is not limited to this. For example, when the sensor unit 140 is a camera, the feature amount may be pixel information.

If the evaluation unit 125 (described later) evaluates the result of the judgment to be an erroneous judgment, the judgment unit 124 at least corrects the judgment result and adjusts the method of judging the object.

For example, the determination unit 124 may determine whether an object is determined to be an obstacle and whether the position of the object moves when the autonomous mobile robot 100 approaches the object, and output the determination result to the evaluation unit 125. Then, when the evaluation unit 125 evaluates that the determination result is an erroneous determination, the determination unit 124 may correct the object from an obstacle to a person for the determination result, and store the corrected determination result as erroneous determination history information in the erroneous determination database 132 described later. In this case, the determination unit 124 may adjust the determination method. For example, in determining whether an object is a person, the determination unit 124 may correct the threshold value so that if the object is determined to be a human leg (i.e., an obstacle) based on a threshold value for determining that the object is a human ankle (i.e., a person) when the diameter of the object is 10 cm or more, the threshold value is determined to be a human leg when the object is a chair leg (i.e., an obstacle). The threshold value may be a range of values with a statistical width, or may be a value obtained by adjusting the standard deviation or sigma.

In addition, when determining whether an object is a person or an obstacle, if the erroneous determination database 132 stores historical information in which the object is erroneously detected as a person and historical information in which the object is erroneously detected as an obstacle, the determination unit 124 may adjust the object determination method so as to determine that the object detected by the sensor unit 140 is an obstacle.

[Evaluation Department]
The evaluation unit 125 evaluates whether the judgment result by the judgment unit is a correct judgment or an incorrect judgment. FIG. 4 is a diagram showing the evaluation of the correctness of the judgment result. For example, as shown in (a) of FIG. 4, when the judgment unit 124 judges an object detected by the sensor unit 140 (in other words, the object to be detected) to be a person when the object is an obstacle, the evaluation unit 125 evaluates that the judgment result by the judgment unit 124 is an incorrect judgment. Also, as shown in (b) of FIG. 4, when the judgment unit 124 judges that the object to be detected is a person when the object is a person, the evaluation unit 125 evaluates that the judgment result by the judgment unit 124 is a correct judgment. Also, as shown in (c) of FIG. 4, when the judgment unit 124 judges that the object to be detected is an obstacle when the object is a person, the evaluation unit 125 evaluates that the judgment result by the judgment unit 124 is an incorrect judgment. Furthermore, as shown in (d) of FIG. 4, when the object to be detected is an obstacle and the determination unit 124 determines that the object is an obstacle, the evaluation unit 125 evaluates that the determination result by the determination unit 124 is a correct determination.

More specifically, when the sensing data of the object acquired by the sensor unit 140 and the judgment result by the judgment unit 124 are registered in the erroneous judgment database 132 as historical information of erroneous judgment, the evaluation unit 125 evaluates that the judgment result by the judgment unit 124 is an erroneous judgment. More specifically, the sensing data of the object includes a feature amount indicating the object, and when the sensor unit 140 includes a position sensor 141 (e.g., LiDAR), the feature amount is, for example, point cloud information indicating the object acquired by the position sensor 141. Also, for example, the feature amount indicating the object included in the sensing data acquired by the camera 142 is pixel information.

The evaluation unit 125 may further evaluate that the judgment result judged by the judgment unit 124 is an erroneous judgment when an input operation to correct the object from a person to an obstacle is received by the reception unit 240 of the terminal device 200 described later, and may register the judgment result as historical information of an erroneous judgment in the erroneous judgment database 132 by linking the sensing data of the object (more specifically, the feature contained in the sensing data), the judgment result, and the corrected judgment result received by the reception unit 240.

In addition, when the determination unit 124 determines that an object is an obstacle and also determines that the position of the object has moved when the autonomous mobile robot 100 approaches the object, the evaluation unit 125 may evaluate the determination result that the object is an obstacle as an erroneous determination, and may link the sensing data of the object (e.g., feature amounts included in the sensing data) and the determination result as historical information of the erroneous determination and register it in the erroneous determination database 132.

Figure 5 is a diagram showing an example of the behavior of the autonomous mobile robot and the person when a person is erroneously determined to be an object. When the autonomous mobile robot 100 determines that the object being detected is an obstacle, it approaches the object and performs an action to avoid the object. At this time, for example, the person may step back (see FIG. 5(a)) or move (see FIG. 5(b)) to avoid a collision with the autonomous mobile robot 100. In this way, when the determination unit 124 determines that an object is an obstacle and the position of the object has moved when the autonomous mobile robot 100 approaches the object, the evaluation unit 125 evaluates that the determination result by the determination unit 124 is an erroneous determination.

[Object position calculation unit]
The object position calculation unit 126 calculates the relative position of the object with respect to the autonomous mobile robot 100 based on the sensing data acquired by the sensor unit 140. For example, if the sensor unit 140 includes a position sensor 141 that can acquire depth data, the object position calculation unit 126 calculates the relative position of the object with respect to the autonomous mobile robot based on the depth data acquired by the position sensor 141.

[Driving plan creation section]
The travel plan creation unit 127 creates a travel plan based on a map showing a specific floor and the autonomous robot's own position. For example, if the autonomous robot 100 is an autonomous robot vacuum cleaner with a cleaning function, the travel plan creation unit 127 may further create a cleaning plan. For example, if there are multiple cleaning areas (e.g., rooms or sections) to be cleaned by the autonomous robot 100, the cleaning plan includes the cleaning order for cleaning these cleaning areas, the travel path and cleaning mode in each cleaning area, etc. The cleaning mode is, for example, a combination of the travel speed of the autonomous robot 100, the suction strength for sucking up dirt on the floor surface, and the rotation speed of the brush.

For example, when an object is detected in the traveling direction (i.e., ahead) of the autonomous mobile robot 100, the driving plan creation unit 127 changes the driving plan based on the map and the self-position, as well as the judgment result by the judgment unit 124 and the relative position of the object with respect to the autonomous mobile robot 100. Specifically, the driving plan creation unit 127 changes the driving plan based on the driving pattern stored in the pattern database 133 described later.

Figure 6 is a diagram showing an example of the operation of the autonomous mobile robot 100 corresponding to the judgment result shown in Figure 4. The travel plan creation unit 127 reflects a travel pattern that avoids an object in the travel plan according to the judgment result by the judgment unit 124. For example, as shown in (a) and (b) of Figure 6, when the judgment unit 124 erroneously judges that an obstacle is a person and when the judgment unit 124 correctly judges that a person is a person, the travel plan creation unit 127 adds a travel pattern in which the autonomous mobile robot 100 avoids an object while keeping a certain distance from the object to the travel plan based on the pattern database 133. Also, for example, as shown in (c) and (d) of Figure 6, when the judgment unit 124 erroneously judges that a person is an obstacle and when the judgment unit 124 correctly judges that an obstacle is an obstacle, the travel plan creation unit 127 adds a travel pattern in which the autonomous mobile robot 100 approaches an object to a predetermined distance and then avoids the object to the travel plan based on the pattern database 133.

When the determination result is corrected by the determination unit 124 as described above with reference to FIG. 5, for example, the driving plan creation unit 127 reflects a driving pattern that avoids objects in accordance with the corrected determination result in the driving plan.

When avoiding a person, the driving plan creation unit 127 may change the cleaning mode, such as slowing down the rotation speed of the brush or stopping the rotation of the brush.

[Drive control unit]
The driving control unit 128 controls the driving unit 160 so that the autonomous mobile robot 100 drives according to a driving plan. More specifically, the driving control unit 128 performs information processing for controlling the operation of the driving unit 160 based on the driving plan. For example, the driving control unit 128 derives control conditions for the driving unit 160 based on information such as a map showing a specific floor and the autonomous robot's own position in addition to the driving plan, and generates a control signal for controlling the operation of the driving unit 160 based on the control conditions. The driving control unit 128 outputs the generated control signal to the driving unit 160.

Details about the derivation of the control conditions for the traveling unit 160 are the same as those for conventional autonomous traveling robots, so a detailed explanation is omitted.

[Cleaning control unit]
The cleaning control unit 129 controls the cleaning unit 170 so that the autonomous mobile robot 100 cleans according to the cleaning plan. More specifically, the cleaning control unit 129 performs information processing for controlling the operation of the cleaning unit 170 based on the cleaning plan. For example, the cleaning control unit 129 derives control conditions for the cleaning unit 170 based on information such as a map showing a specific floor and the self-position in addition to the cleaning plan, and generates a control signal for controlling the operation of the cleaning unit 170 based on the control conditions. The cleaning control unit 129 outputs the generated control signal to the cleaning unit 170. Note that details such as the derivation of the control conditions for the cleaning unit 170 are the same as those of conventional autonomous mobile robot cleaners, and therefore will not be described here.

[Storage]
The storage unit 130 is a storage device that stores a map showing a specific floor, sensing data related to an object detected by the sensor unit 140, and a computer program executed by the control unit 120. The storage unit 130 is realized by, for example, a hard disk drive (HDD), but may also be realized by a semiconductor memory.

The human determination database 131 is a database referenced by the determination unit 124 when determining whether an object is a human or an obstacle. For example, the human determination database 131 stores sensing data (e.g., features indicating a human contained in the sensing data) in advance. For example, the sensing data may be a point cloud, a radius of a circle fitting of the point cloud, and a variance of the points.

The misjudgment database 132 stores history information in which a person was misjudged as an obstacle and history information in which an obstacle was misjudged as a person. The misjudgment database 132 may be a single database, or may be a database divided according to the type of misjudgment. FIG. 7 is a diagram showing an example of data stored in the misjudgment database 132. FIG. 8 is a diagram showing another example of data stored in the misjudgment database 132. For example, as shown in FIG. 7, the misjudgment database 132 may store sensing data, a judgment result, an evaluation result, and a corrected judgment result in association with each other as history information in which a person was misjudged as an obstacle. Also, for example, as shown in FIG. 8, the position where the object was detected, sensing data, a judgment result, an evaluation result, and a corrected judgment result in association with each other as history information in which an obstacle was misjudged as a person may be stored.

The pattern database 133 stores the type of object (person or obstacle), the distance from the autonomous mobile robot 100 to the object, and a driving pattern, which are linked together. Figure 9 is a diagram showing an example of data stored in the pattern database 133. As shown in Figure 9, for each type of object, a driving pattern according to the distance between the autonomous mobile robot 100 and the object is stored.

[Sensor section]
The sensor unit 140 detects objects around the autonomous mobile robot 100 and acquires sensing data related to the detected objects. The sensor unit 140 includes at least one sensor. For example, as shown in Fig. 2, the sensor unit 140 may include a position sensor 141, a camera 142, and an obstacle sensor 143, but this is just an example and is not particularly limited thereto.

For example, the position sensor 141 is installed on the top surface of the autonomous mobile robot 100, and acquires the positional relationship including the distance and direction to the walls around the autonomous mobile robot 100, obstacles in the direction of travel of the autonomous mobile robot 100 (so-called forward), and objects including people. Note that an obstacle is an object other than a wall that divides a specified space, is located in the direction of travel of the autonomous mobile robot 100, and obstructs the travel of the autonomous mobile robot 100. Obstacles are, for example, chairs, desks, beds, sofas, cushions, bags, clothes, portable electrical appliances (for example, irons, humidifiers, air purifiers, etc.). The position sensor 141 may be, for example, a LiDAR or laser range finder that detects the positional relationship (for example, the position of an object and the distance from the robot itself to the object) based on the light that is emitted and reflected by an obstacle and returned.

Camera 142 is, for example, an imaging device that captures images of the surroundings of autonomous mobile robot 100. Camera 142 may capture images or video. Camera 142 may be an RGB camera or an RGB-D camera.

The obstacle sensor 143 detects its surroundings, particularly objects that are obstacles in the traveling direction of the autonomous mobile robot 100 (hereinafter, referred to as obstacle candidates), using a method different from that of the position sensor 141, and acquires the positional relationship of the obstacle candidates with respect to the autonomous mobile robot 100. The obstacle sensor 143 is, for example, an ultrasonic sensor or a distance sensor. For example, the obstacle sensor 143 may be a ToF camera or a stereo camera. In this embodiment, the obstacle sensor 143 is, for example, an ultrasonic sensor. The obstacle sensor 143 has a transmitter 143a arranged in the center of the front of the main body 101, and receivers 143b arranged on both sides of the transmitter 143a, and the receivers 143b receive ultrasonic waves transmitted from the transmitter 143a and reflected by the obstacle candidates, thereby detecting the distance and position of the obstacle candidates.

Although not shown, the sensor unit 140 may include other types of sensors. For example, the other types of sensors may be a floor sensor 144, an encoder, an acceleration sensor, a contact sensor, and the like.

[Notification Department]
The notification unit 150 notifies the user of information. For example, the notification unit 150 may notify the user of the judgment result. For example, the notification unit 150 may notify the user whether the judgment unit 124 judged the object to be a person or an obstacle using at least one of sound, light, and an image. The notification unit 150 may be realized by at least one of a speaker, a lamp, and a display panel, for example. The speaker outputs sound or voice. The lamp lights up or blinks. The display panel is a liquid crystal panel or an organic EL panel, etc., and displays an image. In addition, for example, the notification unit 150 may notify the user of information (for example, the judgment result, etc.) via the communication unit 110. In this case, the notification unit 150 generates information to be notified to the user, and outputs the generated information to the terminal device 200.

[Running section]
The running unit 160 runs the autonomous mobile robot 100 based on instructions from the running control unit 128. The running unit 160 has wheels 161 (see FIG. 3) that run on the floor surface, a running motor (not shown in FIG. 3) that applies torque to the wheels 161, and a housing that accommodates the running motor. The autonomous mobile robot 100 is an opposed two-wheel type equipped with casters 162 (see FIG. 3) as auxiliary wheels, and the two wheels 161 can be independently controlled to allow the autonomous mobile robot 100 to run freely in a straight line, backward, left turn, right turn, and the like. The number of wheels 161 is not limited to two, and may be one, or three or more.

[Cleaning Department]
The cleaning unit 170 sucks dust on the floor surface through suction port 172 (see FIG. 3) based on instructions from the cleaning control unit 129, and stores the sucked dust inside the main body 101 inside the main body 101. The cleaning unit 170 includes a brush rotation motor (not shown in FIG. 3) that rotates the side brushes 171 (see FIG. 2), a suction motor (not shown in FIG. 3) that sucks dust through suction port 172, a power transmission unit (not shown in FIG. 3) that transmits power to these motors, and a storage unit (not shown in FIG. 3) that stores the sucked dust.

[1-2. Terminal device]
Next, the terminal device 200 will be described. The terminal device 200 is, for example, a portable information terminal such as a smartphone or tablet terminal owned by a user, but may also be a stationary information terminal such as a personal computer. The terminal device 200 may also be a dedicated terminal for the autonomous mobile robot system 300. The terminal device 200 includes, for example, a communication unit 210, a control unit 220, a storage unit 230, a reception unit 240, and a display unit 250. Each component will be described below.

[Communications Department]
The communication unit 210 is a communication circuit for the terminal device 200 to communicate with the autonomous mobile robot 100. The communication unit 210 may include a communication circuit (communication module) for communicating via a wide area communication network and a communication circuit (communication module) for communicating via a local communication network. The communication unit 210 is, for example, a wireless communication circuit for performing wireless communication. There is no particular limitation on the communication standard for the communication performed by the communication unit 210.

[Control unit]
The control unit 220 controls the display of images on the reception unit 240, and identifies instructions input by the user (for example, recognizes voice if the input is by voice), etc. The control unit 220 may be realized by, for example, a microcomputer or a processor.

[Storage]
The storage unit 230 is a storage device that stores dedicated application programs and the like to be executed by the control unit 220. The storage unit 230 is realized by, for example, a semiconductor memory.

[Reception department]
The reception unit 240 receives instructions from the user. More specifically, the reception unit 240 receives an input operation performed to transmit the user's instructions to the autonomous mobile robot 100. The reception unit 240 may be realized by, for example, a touch panel, a display panel, a hardware button, or a microphone. The touch panel may be, for example, a capacitive touch panel or a resistive touch panel. The display panel has a function of displaying images and a function of receiving manual input from the user, and receives input operations to a numeric keypad image or the like displayed on a display panel such as a liquid crystal panel or an organic EL (Electro Luminescence) panel. The microphone receives voice input from the user.

[Display]
The display unit 250 displays notifications (e.g., the determination result and the cleaning result) output by the autonomous mobile robot 100, a map of a specific floor, etc. The display unit 250 is, for example, a display panel such as a liquid crystal panel or an organic EL panel.

FIG. 10 is a diagram showing an example of a display screen. As shown in FIG. 10, the display unit 250 displays, for example, a cleaning history map as a cleaning result. The cleaning history map displays, for example, the amount of garbage collected and the determination results of detected objects (for example, an icon showing a person) on a map showing a specific floor cleaned by the autonomous mobile robot 100. The cleaning history map may be generated by the notification unit 150 using map information created by the map creation unit 123 of the autonomous mobile robot 100, or may be created by the map creation unit 123. The display unit 250 may also display an icon for receiving instructions from a user.

In the embodiment, an example has been described in which the autonomous driving robot system 300 includes the autonomous driving robot 100 and the terminal device 200, but the components of the autonomous driving robot system 300 may be distributed in any way among multiple devices. For example, among the components of the autonomous driving robot 100 in the embodiment, the evaluation unit 125 and the erroneous judgment database 132 may be provided in an edge server.

[2. Operation]
Next, an operation of the autonomous mobile robot system 300 according to the embodiment will be described. Figures 11A, 11B, and 11C are flowcharts showing an example of the operation of the autonomous mobile robot system 300 according to the embodiment. In the following, an operation example will be described in which the sensing data acquired by the sensor unit 140 includes point cloud data (point cloud information) acquired by a position sensor 141 (e.g., LiDAR).

First, the autonomous mobile robot 100 is powered on by a user operation (S01). Next, when the control unit 120 of the autonomous mobile robot 100 receives an instruction to start cleaning a specific floor (hereinafter also referred to as a cleaning start instruction) (S02), it causes the acquisition unit 121 to acquire a map showing the specific floor from the memory unit 130 (S03), and causes the sensor unit 140 to start acquiring sensing data (S04). The sensor unit 140 detects objects around the autonomous mobile robot and acquires sensing data related to the detected objects. The sensing data includes, for example, point cloud information indicating the positional relationship between the autonomous mobile robot 100 and the objects.

Next, the acquisition unit 121 of the autonomous mobile robot 100 acquires the sensing data acquired by the sensor unit 140. Here, the acquisition unit 121 acquires, for example, point cloud information indicating the positional relationship of the autonomous mobile robot (i.e., the autonomous mobile robot 100) with surrounding objects (S05).

Next, the self-position calculation unit 122 calculates a self-position indicating the current position of the main body 101 of the autonomous mobile robot 100 on the map based on the acquired map and sensing data (S06).

Next, the driving plan creation unit 127 creates a driving plan based on the map showing the specified floor and the autonomous robot's own position (S07). FIG. 12 is a diagram showing an example of a driving plan. The dashed lines shown in FIG. 12 indicate driving routes. For example, the driving plan creation unit 127 creates a cleaning plan including driving routes and cleaning modes for each room in a specified area. Although not shown in FIG. 11A, when the driving control unit 128 and the cleaning control unit 129 acquire the driving plan created by the driving plan creation unit 127, they start controlling the operation of the driving unit 160 and the cleaning unit 170 according to the acquired driving plan. As a result, the autonomous mobile robot 100 travels and cleans the specified floor (S08).

Next, when the determination unit 124 of the autonomous mobile robot 100 acquires sensing data from the sensor unit 140 (not shown), it determines whether an object has been detected in the traveling direction of the autonomous mobile robot 100, i.e., ahead (S09).

If the determination unit 124 determines that no object has been detected in front of the autonomous mobile robot 100 (No in S09), the control unit 120 determines whether or not cleaning has been completed (S10). If the control unit 120 determines that cleaning has not been completed (No in S10), the process returns to step S05. On the other hand, if the control unit 120 determines that cleaning has been completed (Yes in S10), it transmits the cleaning results (e.g., a cleaning history map) to the terminal device 200, and when the terminal device 200 obtains the cleaning results via the communication unit 210, it causes the display unit 250 to display the cleaning results. As a result, the display unit 250 of the terminal device 200 presents the cleaning results to the user (S11).

Next, when the input operation of an instruction to end cleaning (hereinafter, an end instruction) is received by the reception unit 240 of the terminal device 200 (S12), the control unit 120 ends the cleaning.

On the other hand, in step S09, if the determination unit 124 determines that an object has been detected in front of the autonomous mobile robot 100 (Yes in S09), it extracts features (here, a point cloud) from the sensing data (not shown), and determines whether the extracted point cloud represents a person (S13). For example, the determination unit 124 determines whether the point cloud represents a person based on the person determination database 131.

If the determination unit 124 determines that the point cloud does not represent a person (No in S13), it determines that the object detected in front of the autonomous mobile robot 100 is an obstacle (not shown). Next, the evaluation unit 125 determines whether or not the point cloud data (hereinafter, point cloud information) is stored in the erroneous judgment database 132 (S14), and if it determines that the point cloud information is not stored in the erroneous judgment database 132 (No in S14), it evaluates the judgment result by the determination unit 124 as a correct judgment (not shown).

Although not shown, the travel plan creation unit 127 reflects a travel pattern for avoiding the obstacle in the travel plan based on the determination result by the determination unit 124, and the travel control unit 128 controls the operation of the travel unit 160 according to the travel plan. As a result, the autonomous mobile robot 100 approaches the object according to the travel pattern reflected in the travel plan by the travel plan creation unit 127 (S15). More specifically, a description will be given with reference to FIG. 13. FIG. 13 is a diagram showing the operation when the autonomous mobile robot 100 determines that the detected object is an obstacle. As shown in FIG. 13, the autonomous mobile robot 100 approaches the obstacle 1 according to the travel plan. At this time, the determination unit 124 determines whether the position of the object has moved (S16). Then, when the determination unit 124 determines that the position of the object has not moved (No in S16), the evaluation unit 125 evaluates that the determination result by the determination unit 124 that the object is an obstacle is a correct determination (S17).

Next, the object position calculation unit 126 calculates the position of the object based on the self-position and the positional relationship (S18). Then, the driving control unit 128 controls the operation of the driving unit 160 according to the driving plan, so that the autonomous mobile robot 100 approaches the object (here, an obstacle) and avoids it (S19). More specifically, this will be described with reference to FIG. 14. FIG. 14 is a diagram showing a schematic diagram of the operation of the autonomous mobile robot 100 to avoid an obstacle. The autonomous mobile robot 100 approaches the obstacle 1, stops in front of the obstacle 1 (for example, at a distance of 100 mm), turns around, and continues driving.

On the other hand, in step S14, if the evaluation unit 125 determines that the point cloud information is stored in the erroneous judgment database 132 (Yes in S14), the evaluation unit 125 evaluates that the judgment result by the judgment unit 124 is an erroneous judgment (S20). In addition, in step S16, if the judgment unit 124 determines that the position of the object has moved when the autonomous mobile robot 100 approaches the object in step S15 (Yes in S16), the evaluation unit 125 evaluates that the judgment result by the judgment unit 124 is an erroneous judgment (S20). Specifically, the description will be given with reference to FIG. 15. FIG. 15 is a diagram that shows the movement of the person 2 when the autonomous mobile robot 100 approaches the person 2. As shown in FIG. 15, when the judgment unit 124 determines that the object is an obstacle, the autonomous mobile robot 100 approaches the object, but when the object is not an obstacle but a person, the person 2 moves backward (arrow A) to avoid the autonomous mobile robot 100. At this time, the determination unit 124 of the autonomous mobile robot 100 determines that the object has moved, and the evaluation unit 125 evaluates that the determination result that the object is an obstacle based on the determination of the object's movement by the determination unit 124 is an erroneous determination.

The evaluation unit 125 may evaluate that the determination result that the object is an obstacle is an erroneous determination by detecting that the autonomous mobile robot 100 has come into contact with a person using the sensor unit 140 (e.g., a contact sensor). Figures 16 and 17 are diagrams that show the operation of the autonomous mobile robot 100 when it has erroneously determined that the object is an obstacle. When the determination unit 124 determines that the object is an obstacle, the autonomous mobile robot 100 approaches the object, but when the person 2 does not notice the approach of the autonomous mobile robot 100, the autonomous mobile robot 100 comes into contact with the person 2.

Next, when the evaluation unit 125 evaluates in step S20 that the judgment result that the object is an obstacle is an incorrect judgment, the judgment unit 124 corrects the judgment result, more specifically, for the judgment result evaluated in step S20 as an incorrect judgment, the judgment unit 124 corrects the object from an obstacle to a person (S21).

Next, the object position calculation unit 126 calculates the position of the object based on the self-position and the positional relationship (S22). Then, the driving control unit 128 controls the operation of the driving unit 160 according to the driving plan, so that the autonomous mobile robot 100 keeps a distance from the object (here, a person) and largely avoids the object (S23). More specifically, this will be described with reference to FIGS. 18 and 19. FIG. 18 is a diagram that shows a schematic diagram of the operation of the autonomous mobile robot 100 when it determines that the object it has detected is a person. FIG. 19 is a diagram that shows a schematic diagram of the operation of the autonomous mobile robot 100 to avoid the person 2. As shown in FIGS. 18 and 19, the autonomous mobile robot 100 keeps a predetermined distance from the person 2 and largely avoids the area around the person.

On the other hand, in step S13, if the determination unit 124 determines that the point cloud indicates a person (Yes in S13), it determines that the object detected in front of the autonomous mobile robot 100 is a person (not shown). Next, the evaluation unit 125 determines whether the point cloud data (so-called point cloud information) is stored in the erroneous judgment database 132 (S24), and if it determines that the point cloud information is not stored in the erroneous judgment database 132 (No in S14), it performs the processing of step S22. On the other hand, if the evaluation unit 125 determines that the point cloud information is stored in the erroneous judgment database 132 (Yes in S24), it evaluates the determination result by the determination unit 124 as an erroneous judgment (S25).

Next, if the evaluation unit 125 evaluates in step S25 that the judgment result that the object is a person is an erroneous judgment, the judgment unit 124 corrects the judgment result; more specifically, for the judgment result evaluated in step S25 as an erroneous judgment, the judgment unit 124 corrects the object from a person to an obstacle (S26), and then performs the process of step S18.

Next, the control unit 120 of the autonomous mobile robot 100 determines whether or not cleaning is complete after processing steps S19 and S23 (S27). If the control unit 120 determines that cleaning is not complete (No in S27), the process returns to step S05. On the other hand, if the control unit 120 determines that cleaning is complete (Yes in S27), the control unit 120 determines whether or not people were avoided during cleaning (S28). If the control unit 120 determines that people were not avoided during cleaning (No in S28), the process proceeds to step S11.

On the other hand, in step S28, if the control unit 120 determines that a person is being avoided during cleaning (Yes in S28), it identifies the position where the person was avoided during cleaning (hereinafter also referred to as the person avoidance position) and outputs the position information to the driving plan creation unit 127 (not shown). The driving plan creation unit 127 creates a driving plan to revisit the position where the person was avoided during cleaning based on the map, the self-position, and the position information, and outputs the created driving plan to the driving control unit 128 (not shown). The operation of the driving unit 160 is controlled by the driving control unit 128, so that the autonomous mobile robot 100 revisits the position where the person was avoided (S29). More specifically, a description will be given with reference to FIG. 20. FIG. 20 is a diagram that illustrates the operation of the autonomous mobile robot revisiting the position where the person was avoided. As shown in FIG. 20, the autonomous mobile robot 100 revisits the position where the person was avoided, and determines whether an object is detected again at the position where the object was detected during cleaning, and if an object is detected, whether it is a person or not. Although not shown in FIG. 11C and FIG. 20, if no object is detected by the sensor unit 140, the autonomous mobile robot 100 cleans a position that avoids the person 2.

When the determination unit 124 detects an object again at a position where a person was avoided and determines that the object is a person (Yes in S30), the control unit 120 causes the map creation unit 123 to link the detected position to the cleaning history map (denoted as map in FIG. 11C) (S31). Next, the control unit 120 determines whether or not all person avoidance positions have been revisited (S32). On the other hand, when the determination unit 124 determines that an object has not been detected again at a position where a person was avoided (No in S30), the process of step S32 is performed.

When the control unit 120 determines that all human avoidance positions have not been revisited (No in S32), the process returns to step S29. On the other hand, when the control unit 120 determines that all human avoidance positions have been revisited (Yes in S32), the control unit 120 outputs the cleaning results (e.g., a cleaning history map) to the terminal device 200 (not shown). When the control unit 220 of the terminal device 200 acquires the cleaning results output from the autonomous mobile robot 100, the control unit 220 presents the cleaning results to the user by displaying them on the display unit 250 (S33). Here, the process of step S33 will be specifically described with reference to FIG. 22.

21 is a diagram showing an example of the cleaning result displayed on the display screen. For example, FIG. 21 shows the history details of the cleaning history map shown in FIG. 10. In the process of step S33, when the control unit 220 of the terminal device 200 acquires the cleaning result (see FIG. 10) output from the autonomous mobile robot 100, the control unit 220 displays the cleaning result on the display unit 250. Then, for example, when the user touches a person mark on the cleaning history map displayed on the display unit 250 in FIG. 10, the reception unit 240 receives an instruction to enlarge and display the area around the person mark. Then, based on the instruction, the control unit 220 causes the display unit 250 to enlarge and display the map around the person mark. Furthermore, when the person touches the history details tab, the reception unit 240 receives an instruction to display the travel trajectory of the autonomous mobile robot 100. Then, the control unit 220 causes the display unit 250 to display the cleaning result shown in FIG. 21. In this way, the process of step S33 may include a process of receiving an instruction regarding the display and changing the display.

Next, when the reception unit 240 receives an instruction to correct the judgment result (hereinafter, also referred to as a correction instruction) (S34), the control unit 220 of the terminal device 200 outputs the correction instruction to the autonomous mobile robot 100 (not shown in FIG. 11C). When the control unit 120 of the autonomous mobile robot 100 acquires the correction instruction (not shown), the evaluation unit 125 updates the erroneous judgment database 132 by linking the sensing data, the judgment result, and the corrected judgment result and storing them in the erroneous judgment database 132 (S35). FIG. 22 is a diagram showing an example of accepting an input operation to correct the judgment result. As shown in FIG. 22, for example, when the user touches the person mark, the reception unit 240 accepts the correction request, and the control unit 220 causes the display unit 250 to display a "message on whether or not to change the person to an obstacle." When the user touches "Yes", the reception unit 240 receives a correction instruction to change the person to an obstacle, and the control unit 220 outputs the correction instruction to the autonomous mobile robot 100.

When the control unit 220 receives an instruction to end the correction of the judgment result from the reception unit 240 (S36), the control unit 220 outputs the instruction to the autonomous mobile robot 100 (not shown). Then, when the control unit 120 of the autonomous mobile robot 100 receives the instruction, it outputs the corrected cleaning result (e.g., a history map of the corrected cleaning) to the terminal device 200 (not shown). Then, when the control unit 220 of the terminal device 200 receives the corrected cleaning result, it causes the display unit 250 to display the corrected cleaning result. Note that the process of step S36 may be performed after the corrected cleaning result is presented to the user. In other words, after the user checks the corrected cleaning result, the reception unit 240 may accept the end instruction by, for example, touching an end button displayed on the display unit 250.

[3. Effects, etc.]
As described above, the autonomous driving robot system 300 according to the embodiment is an autonomous driving robot system equipped with an autonomous driving robot 100 that autonomously drives on a predetermined floor, and is equipped with a sensor unit 140 that detects objects around the autonomous driving robot 100 and acquires sensing data related to the detected objects, a judgment unit 124 that judges whether the object is a person or an obstacle based on the sensing data, and an evaluation unit 125 that evaluates the judgment result by the judgment unit 124 to be an erroneous judgment when the sensing data and the judgment result are stored in the erroneous judgment database 132 as historical information of an erroneous judgment, and when the judgment unit 124 evaluates that the judgment result is an erroneous judgment by the evaluation unit 125, the judgment unit 124 at least either corrects the judgment result or adjusts the method of judging the object.

As a result, when the autonomous mobile robot system 300 judges that the result of its judgment as to whether an object is a person or an obstacle is an erroneous judgment, it at least corrects the judgment result and adjusts the judgment method, making it less likely to repeat the same erroneous judgment.

For example, in the autonomous mobile robot system 300 according to the embodiment, the judgment unit 124 extracts features included in the sensing data and judges whether an object is a human or an obstacle based on the extracted features. The erroneous judgment database 132 stores the results of judgment in which an object is judged to be an obstacle when it is a human, and the results of judgment in which an object is judged to be an obstacle when it is a human, as historical information of erroneous judgment, in association with the features included in the sensing data, the judgment results, and the corrected judgment results.

This allows the autonomous mobile robot system 300 to determine whether an object is a person or an obstacle based on the feature amounts contained in the sensing data. The feature amounts may be, for example, a point cloud detected by LiDAR, or pixel information contained in an image captured by a camera. Furthermore, the autonomous mobile robot system 300 can evaluate the accuracy of the determination result based on the erroneous determination database 132, which stores historical information on erroneous determinations.

For example, the autonomous mobile robot system 300 according to the embodiment further includes a notification unit 150 that notifies the user of information and a reception unit 240 that receives an input operation by the user, the notification unit 150 notifies the user of the judgment result, the reception unit 240 receives an input operation by the user to correct the judgment result, and the evaluation unit 125 further evaluates the judgment result as an erroneous judgment when the reception unit 240 receives an input operation to correct the object from a person to an obstacle in the judgment result, and stores the judgment result as historical information of the erroneous judgment in the erroneous judgment database 132 by linking the feature included in the sensing data, the judgment result, and the corrected judgment result.

By accepting an input operation by the user to correct the judgment result, the autonomous mobile robot system 300 stores the judgment result in which an obstacle is erroneously judged to be a person in the erroneous judgment database 132 as historical information of the erroneous judgment. Therefore, by judging an object based on the erroneous judgment database 132, the autonomous mobile robot system 300 can reduce the probability of erroneously judging a similar obstacle to be a person.

For example, in the autonomous mobile robot system 300 according to the embodiment, when the determination unit 124 determines that an object is an obstacle, the autonomous mobile robot 100 approaches the object, and when the determination unit 124 determines that the position of the object has moved when the autonomous mobile robot 100 approaches the object, the evaluation unit 125 evaluates the determination result that the object is an obstacle to be an erroneous determination, and links the feature amount included in the sensing data related to the object with the determination result as historical information of the erroneous determination and stores the result in the erroneous determination database, and the determination unit 124 corrects the object from an obstacle to a person for the determination result evaluated as an erroneous determination, and stores the corrected determination result in the erroneous determination database 132 as historical information of the erroneous determination.

As a result, for example, when the determination unit 124 determines that an object is an obstacle and the determination result is not stored in the erroneous determination database 132, the autonomous mobile robot system 300 can evaluate whether or not the determination is an erroneous determination based on whether or not the object moves when the autonomous mobile robot system 300 approaches the object. Furthermore, when the determination is evaluated as an erroneous determination, the autonomous mobile robot system 300 updates the erroneous determination database 132, thereby reducing the probability of again erroneously determining that a person is an obstacle.

For example, in the autonomous mobile robot system 300 according to the embodiment, when determining whether an object is a person or an obstacle, if the erroneous judgment database 132 stores historical information in which the object is erroneously detected as a person and historical information in which the object is erroneously detected as an obstacle, the judgment unit 124 adjusts the object judgment method so as to judge the object as an obstacle.

As a result, the autonomous mobile robot system 300 can adjust the determination method so that an object is determined to be an obstacle based on the sensing data, thereby enabling the autonomous mobile robot 100 to avoid colliding with a person.

For example, in the autonomous robot system 300 according to the embodiment, the autonomous robot 100 is an autonomous robot vacuum cleaner with a cleaning function.

As a result, the autonomous mobile robot system 300 allows the autonomous mobile robot 100 to safely navigate and clean a specified space.

The judgment adjustment method according to the embodiment is a judgment adjustment method executed by one or more computers, which detect objects around the autonomous mobile robot 100, acquire sensing data related to the detected objects, and determine whether the object is a person or an obstacle based on the acquired sensing data. If the sensing data and the judgment result are registered in the erroneous judgment database 132 as historical information of an erroneous judgment, the judgment result is evaluated as an erroneous judgment. If the judgment result is evaluated as an erroneous judgment, the judgment result is corrected and/or the object judgment method is adjusted.

As a result, if the determination result of whether an object is a person or an obstacle is evaluated as an erroneous determination, the determination method at least corrects the determination result and adjusts the determination method, making it less likely that a similar erroneous determination will be repeated.

Other Embodiments
The autonomous mobile robot system and the judgment adjustment method according to one or more aspects of the present disclosure have been described based on the above-mentioned embodiments, but the present disclosure is not limited to these embodiments. As long as they do not deviate from the gist of the present disclosure, various modifications conceived by a person skilled in the art and forms formed by combining components of different embodiments may also be included within the scope of one or more aspects of the present disclosure.

Furthermore, each processing unit included in the autonomous mobile robot system according to the above embodiment is realized as an LSI, which is typically an integrated circuit. These may be individually implemented as single chips, or may be integrated into a single chip that includes some or all of them.

In addition, the integrated circuit is not limited to LSI, but may be realized by a dedicated circuit or a general-purpose processor. It is also possible to use an FPGA (Field Programmable Gate Array) that can be programmed after LSI manufacturing, or a reconfigurable processor that can reconfigure the connections and settings of circuit cells inside the LSI.

In the above embodiment, each component may be configured with dedicated hardware, or may be realized by executing a software program suitable for each component. Each component may be realized by a program execution unit such as a CPU or processor reading and executing a software program recorded on a storage medium such as a hard disk or semiconductor memory.

Furthermore, all the numbers used above are examples to specifically explain this disclosure, and the embodiments of this disclosure are not limited to the numbers exemplified.

The division of functional blocks in the block diagram is one example, and multiple functional blocks may be realized as one functional block, one functional block may be divided into multiple blocks, or some functions may be transferred to other functional blocks. In addition, the functions of multiple functional blocks having similar functions may be processed in parallel or in a time-sharing manner by a single piece of hardware or software.

The order in which each step in the flowchart is performed is merely an example to specifically explain the present disclosure, and orders other than those described above may also be used. Some of the steps may also be performed simultaneously (in parallel) with other steps.

This disclosure is widely applicable to autonomous robot systems equipped with autonomous robots.

10 Wide area communication network 100 Autonomous mobile robot 101 Main body 110 Communication unit 120 Control unit 121 Acquisition unit 122 Self-position calculation unit 123 Map creation unit 124 Determination unit 125 Evaluation unit 126 Object position calculation unit 127 Travel plan creation unit 128 Travel control unit 129 Cleaning control unit 130 Memory unit 131 Person determination database 132 False determination database 133 Pattern database 140 Sensor unit 141 Position sensor 142 Camera 143 Obstacle sensor 143a Transmission unit 143b Reception unit 144 Floor sensor 150 Notification unit 160 Travel unit 161 Wheels 162 Casters 170 Cleaning unit 171 Side brush 172 Suction port 200 Terminal device 210 Communication unit 220 Control unit 230 Storage unit 240 Reception unit 250 Display unit 300 Autonomous traveling robot system

Claims (8)

An autonomous mobile robot system including an autonomous mobile robot that autonomously travels on a predetermined floor,
A sensor unit that detects objects around the autonomous robot and acquires sensing data relating to the detected objects;
a determination unit that determines whether the object is a person or an obstacle based on the sensing data;
an evaluation unit that evaluates the determination result by the determination unit as an erroneous determination when the sensing data and the determination result are stored in an erroneous determination database as history information of erroneous determination;
Equipped with
When the evaluation unit evaluates that the determination result is an erroneous determination, the determination unit at least one of correcting the determination result and adjusting a method for determining the object.
Autonomous driving robot system. The determination unit extracts a feature amount included in the sensing data, and determines whether the object is a person or an obstacle based on the extracted feature amount;
The erroneous determination database stores, as history information of the erroneous determination, a determination result in which the object is determined to be an obstacle when the object is a person, and a determination result in which the object is determined to be an obstacle when the object is a person, in association with the feature amount included in the sensing data, the determination result, and a corrected determination result.
The autonomous mobile robot system according to claim 1 . Furthermore, a notification unit that notifies the user of information;
A reception unit that receives an input operation by the user;
Equipped with
The notification unit notifies the user of the determination result,
The accepting unit accepts an input operation of a correction to the determination result by the user,
The evaluation unit further
When the receiving unit receives an input operation for correcting the object from a person to an obstacle in response to the determination result, the determination result is evaluated as an erroneous determination,
the determination result is stored as the history information of the misjudgment in the misjudgment database in association with the feature amount included in the sensing data, the determination result, and a corrected determination result;
The autonomous mobile robot system according to claim 2 . When the determination unit determines that the object is an obstacle, the autonomous mobile robot approaches the object,
The evaluation unit is
When the determination unit determines that the position of the object has moved when the autonomous mobile robot approaches the object, the determination result that the object is an obstacle is evaluated as an erroneous determination,
storing, as the history information of the erroneous determination, the feature amount included in the sensing data regarding the object and the determination result in the erroneous determination database in association with each other;
The determination unit is
When the determination result is evaluated as being an erroneous determination, the object is corrected from an obstacle to a person;
storing the corrected judgment result in the erroneous judgment database as the history information of the erroneous judgment;
4. The autonomous mobile robot system according to claim 2 or 3. the determination unit adjusts a method of determining whether the object is a person or an obstacle, when history information in which the object is erroneously detected as a person and history information in which the object is erroneously detected as an obstacle are stored in the erroneous determination database, so as to determine the object as an obstacle.
5. The autonomous mobile robot system according to claim 1 . The autonomous robot is an autonomous robot cleaner having a cleaning function.
6. The autonomous mobile robot system according to claim 1 . 1. A method for adjusting decisions executed by one or more computers, comprising:
The one or more computers
Detecting an object around the autonomous robot and acquiring sensing data relating to the detected object;
determining whether the object is a person or an obstacle based on the acquired sensing data;
If the sensing data and the judgment result are registered in an erroneous judgment database as history information of an erroneous judgment, the judgment result is evaluated as an erroneous judgment;
When the determination result is evaluated as an erroneous determination, at least one of correcting the determination result and adjusting the method for determining the object is performed.
Judgment adjustment method. A program for causing one or more computers to execute the judgment adjustment method according to claim 7.

Images