A system for estimating the location of workers.
The system addresses drift in motion sensors by associating worker movements with pre-defined tasks and locations, using a reference position to reset the estimated position at each task, thereby reducing errors in position estimation.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- TOYOTA JIDOSHA KK
- Filing Date
- 2024-10-29
- Publication Date
- 2026-05-15
Smart Images
Figure 2026078681000001_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to a system for estimating the position of an operator.
Background Art
[0002] Various techniques for estimating the work being performed by an operator and the position of the operator using sensors have been proposed. For example, Patent Document 1 discloses a technique for estimating the work and position of an operator using an acceleration sensor and an angular velocity sensor.
Prior Art Documents
Patent Documents
[0003]
Patent Document 1
Summary of the Invention
Problems to be Solved by the Invention
[0004] In motion sensors including acceleration sensors and angular velocity sensors, a phenomenon called drift is known. Drift is a phenomenon in which the error of the sensor accumulates over time. Therefore, when using a motion sensor to estimate the position of an operator, there is a risk that the deviation of the estimated position information will increase over time.
Means for Solving the Problems
[0005] The present disclosure can be realized in the following forms.
[0006] (1) According to one embodiment of the present disclosure, a system for estimating the position of a worker is provided. The system comprises a motion sensor attached to a worker and detecting the worker's movements; a work estimation unit that estimates the work being performed by the worker using the detected movements; and a position estimation unit that estimates the position of the worker using the estimated work, and further estimates that the worker is located at a location pre-associated with the estimated work. In this type of system, the position estimation unit estimates that the worker is located at a location pre-associated with the estimated task. Therefore, each time the task is performed, the worker is estimated to be located at the pre-associated location. Compared to a configuration in which the position estimation unit does not perform such position estimation, this configuration can suppress the increase in the deviation of the worker's estimated position over time due to motion sensor drift. (2) In the system described above, the operation may be performed repeatedly at predetermined time intervals. In this type of system, since the work is performed repeatedly at predetermined time intervals, it is presumed that the worker is located at a pre-associated location at each of these time intervals. This helps to further suppress the increase in the estimated position of the worker over time. (3) In the system of the above form, the position estimation unit may, after estimating that the worker is located at the pre-associated location, set that location as the reference position, estimate the amount of displacement of the worker using the motion sensor, and estimate the position of the worker using the reference position and the amount of displacement. In this type of system, the position estimation unit first estimates that the worker is located at a pre-associated location, sets that location as the reference position, estimates the displacement using a motion sensor, and then estimates the worker's position using the reference position and the displacement. Therefore, the reference position can be set each time work is estimated, and the worker's position can be estimated using the displacement from that reference position. This makes it possible to further suppress the increase in the estimated worker's position deviation over time. (4) In the system of the above form, the motion sensor may include an inertial measuring device attached to the worker. In this type of system, the motion sensor includes an inertial measuring device, which allows for the detection of large movements, including changes in the worker's skeletal structure. (5) In the system of the above form, the operation includes the movement of the worker's fingers, and the motion sensor may be attached to the fingers. In this type of system, the motion sensor is attached to the fingers, allowing for the detection of fine movements of the worker's fingers, and the work estimation unit can estimate tasks that involve fine movements of the fingers. [Brief explanation of the drawing]
[0007] [Figure 1] This figure shows a schematic configuration of a system for estimating the position of a worker in one embodiment of the present disclosure. [Figure 2] This diagram illustrates the use of a system to estimate the position of a worker. [Figure 3] This is a flowchart showing the procedure for estimating the worker's position. [Figure 4] This figure illustrates a comparison of the cumulative errors of position estimation between the examples and comparative examples. [Modes for carrying out the invention]
[0008] A. Embodiments: A1. Configuration of System 100: Figure 1 shows a schematic configuration of a system 100 for estimating the position of a worker WK in one embodiment of the present disclosure (hereinafter also referred to as the "system"). The system 100 is used to estimate the position of a worker WK performing a task. The system 100 is used in a workplace where the worker WK performs a task. In this embodiment, the workplace is a factory. The system 100 comprises a motion sensor 200 and a position estimation device 300.
[0009] <Configuration of motion sensor 200> The motion sensor 200 detects the movements of worker WK. The motion sensor 200 is attached to worker WK. In this embodiment, the motion sensor 200 is attached to the head, left and right arms, chest, abdomen, and left and right legs of worker WK. The motion sensor 200 is also attached along the skeleton of worker WK. The motion sensor 200 is configured to detect physical quantities related to changes in the movement of each part of worker WK. More specifically, the motion sensor 200 detects acceleration and angular velocity occurring in each part of worker WK as physical quantities related to changes in motion. In this embodiment, the motion sensor 200 is configured as an inertial measurement unit (IMU) equipped with a 3-axis acceleration sensor, a 3-axis gyroscope sensor, and a 3-axis geomagnetic sensor. The motion sensor 200 detects the movements of worker WK using the detected acceleration and angular velocity. The detected movements are transmitted to the position estimation device 300.
[0010] <Configuration of position estimation device 300> The position estimation device 300 is composed of a computer comprising a processor 110, memory 120, input / output interface 130, and internal bus 140. The processor 110 is, for example, a CPU (Central Processing Unit). The memory 120 is, for example, an HDD (Hard Disk Drive) or SSD (Solid State Drive). A communication device 150 and a display device 160 are connected to the input / output interface 130. The communication device 150 can communicate with the motion sensor 200 by wired or wireless communication. The display device 160 is, for example, a liquid crystal display and displays various information in the system 100. The internal bus 140 connects the processor 110, the memory 120, and the input / output interface 130 so that they can communicate with each other. The processor 110 makes the work estimation unit 111 and the position estimation unit 112 function by executing the program 121 stored in the memory 120.
[0011] The work estimation unit 111 estimates the work being performed by worker WK. Specifically, the work estimation unit 111 estimates the work using the movements of worker WK detected by the motion sensor 200 and the work information stored in the work memory unit 122 of the memory 120. The work information is pre-stored with movements and tasks linked together. For example, the work memory unit 122 stores work information linked to the movement of raising both arms to chest height and the task of picking up the workpiece. In other words, characteristic movements in a task are stored linked to that task. The linking of movements and tasks is performed, for example, by using machine learning. The work estimation unit 111 determines whether the movement detected by the motion sensor 200 is stored as work information, and if it is stored as work information, it estimates that worker WK is performing the task stored linked to that movement.
[0012] The position estimation unit 112 estimates the position of worker WK. Specifically, the position estimation unit 112 estimates the position of worker WK using a reference position and a displacement. The reference position will be described later. The displacement is obtained by calculating the motion acceleration of worker WK using the acceleration and motion detected by the motion sensor 200, and then performing a double integral of the calculated motion acceleration. Hereafter, the position estimation using the displacement from the reference position will also be referred to as the "first estimation process".
[0013] Furthermore, the position estimation unit 112 in this disclosure estimates that worker WK is located at a location pre-associated with the work estimated by the work estimation unit 111. Specifically, the position estimation unit 112 estimates the position of worker WK using the estimated work and the work location information stored in the position storage unit 123 of the memory 120. The work location information is pre-stored with work and location associated. Within the factory, a given work is performed at a given location. Therefore, the position of worker WK can be estimated from the work that worker WK is performing. As work location information, for example, the work of removing a workpiece and the location where the workpiece is placed are stored as associated information. Therefore, the position estimation unit 112 determines whether the work estimated by the work estimation unit 111 is stored as work location information, and if the work is stored as work location information, it estimates that worker WK is located at the location associated with that work. In the following, the position estimation process using estimated work and work location information will also be referred to as the "second estimation process."
[0014] The position estimation unit 112 sets the reference position in the first estimation process described above to the position estimated by the second estimation process. That is, the position estimation unit 112 executes the second estimation process when a task stored in the work position information is performed by worker WK, and estimates that worker WK is located at the position associated with that task. Subsequently, as the first estimation process, the position estimation unit 112 sets the position of worker WK estimated by the second estimation process as the reference position, and estimates the position of worker WK using the displacement amount from the reference position. That is, if a task not stored in the work position information is performed by worker WK, or if worker WK is moving between tasks, the first estimation process is executed using the position estimated by the immediately preceding second estimation process as the reference position.
[0015] In cases where the preceding second estimation process has not been performed, i.e., when the process is performed for the first time after the system 100 is started, the reference position is set to, for example, the position of worker WK at the time the system 100 is started.
[0016] A2. An example of estimating the position of operator WK: FIG. 2 is a diagram for explaining the estimation of the position of operator WK using system 100. In the example shown in FIG. 2, operator WK performs each of the three operations ST1, ST2, and ST3 at positions P1, P2, and P3, respectively. Also, the three operations ST1, ST2, and ST3 are repeatedly executed in this order. Operation ST1 is an operation of lifting work W1. Work W1 is placed at position P1. Operation ST2 is an operation of bolting work W1 using driver DR. Driver DR and workbench WB are placed at position P2. Operation ST3 is an operation of placing the bolted work W1 at work storage SP. Work storage SP is at position P3.
[0017] Work memory unit 122 pre-stores by associating the motions detected by motion sensor 200 with each of operations ST1, ST2, and ST3. Specifically, the motion of operator WK raising both arms is associated with operation ST1 of lifting work W1, the motion of operator WK using driver DR is associated with operation ST2 of bolting work W1, and the motion of operator WK lowering both hands is associated with the operation of placing work W1.
[0018] Position memory unit 123 pre-stores by associating operations with positions. Specifically, operation ST1 is associated with position P1, operation ST2 is associated with position P2, and operation ST3 is associated with position P3.
[0019] Work estimation unit 111 estimates the work of operator WK using the motion detected by motion sensor 200 and work memory unit 122. For example, when motion sensor 200 detects that operator WK raises both arms, work estimation unit 111 estimates that operation ST1 of lifting work W1 is being performed.
[0020] The position estimation unit 112 uses the estimated task and the position storage unit 123 to estimate the position of worker WK. In other words, position estimation is performed by the second estimation process. For example, if the task estimation unit 111 estimates that worker WK is performing task ST1, which involves lifting workpiece W1, the position estimation unit 112 estimates that worker WK is located at position P1.
[0021] Next, worker WK moves from position P1 to position P2 in order to perform task ST2. Position estimation during this movement is performed by a second estimation process that uses position P1 as the reference position. That is, position estimation is performed by calculating the displacement of worker WK from position P1 using the motion acceleration detected by the motion sensor 200. This position estimation is repeatedly performed while worker WK repeatedly performs tasks ST1 to ST3.
[0022] A3. Method for estimating the location of worker WK: Figure 3 is a flowchart illustrating the procedure for estimating the location of worker WK. This method is used, for example, for analyzing worker WK in a factory. This method is executed when worker WK is performing work. Prior to the execution of this method, work information, which links worker WK's actions to their tasks, and work location information, which links their tasks to their locations, are prepared in advance.
[0023] The work estimation unit 111 uses the movements and work information of worker WK detected by the motion sensor 200 to estimate the work being performed by worker WK (S100). The position estimation unit 112 uses the estimated work and work position information to estimate that worker WK is at a predetermined position (S110). The predetermined position refers to a position stored as work position information, associated with the work. In other words, in S110, the position estimation unit 112 performs a second estimation process.
[0024] Following S110, a first estimation process may be performed using the estimated position of worker WK as the reference position.
[0025] This method may be repeated while worker WK is performing the work. Alternatively, this method may be repeated at predetermined time intervals.
[0026] A4. Comparison of Examples and Comparative Examples: Figure 4 is a diagram illustrating a comparison of the cumulative error of position estimation between the example and the comparative example. In Figure 4, the solid line represents the example, and the dashed line represents the comparative example. In the example and the comparative example, the cumulative error of position estimation when n tasks 1 to n are performed in sequence was examined. Figure 4 schematically represents the cumulative error. In the example, the position of worker WK was estimated using the first estimation process and the second estimation process. In the example, the actions associated with each of tasks 1 to n are stored as work information in the work memory unit 122, and the positions associated with each of tasks 1 to n are stored as work position information in the position memory unit 123. That is, when it is estimated that tasks 1 to n have been performed, it is estimated that worker WK is located at the position associated with each of tasks 1 to n. Furthermore, between each of tasks 1 to n, the position of worker WK is estimated using the displacement amount from the position that was immediately estimated as the position of worker WK and is associated with tasks 1 to n, which is used as the reference position.
[0027] In contrast, in the comparative example, position estimation was performed using only the first estimation process. Specifically, the position of worker WK was estimated using only the displacement amount of worker WK calculated by the motion sensor 200. The reference position for the displacement amount is the starting point of work 1.
[0028] In the embodiment shown by the solid line in Figure 4, each time the work estimation unit 111 estimates that worker WK is performing work 1 to n, the position estimation unit 112 estimates that worker WK is at the position associated with work 1 to n. Since the position estimation accuracy at this time is relatively high, the error in position information is almost zero. Therefore, each time it is estimated that worker WK is performing work 1 to n, the cumulative error in position information is reset. In contrast, in the comparative example shown by the dashed line in Figure 4, the same processing as in the embodiment is not performed, so the cumulative error in position information increases over time.
[0029] As is clear from the comparison between the example and the comparative example, in the example, the cumulative error is reset each time it is estimated that worker WK is performing tasks 1 to n, thus suppressing the increase in the cumulative error over time.
[0030] According to the system 100 of the embodiment described above, the position estimation unit 112 estimates that worker WK is located at a location pre-associated with the estimated task. Therefore, each time the task is performed, worker WK is estimated to be located at the pre-associated location. Compared to a configuration in which the position estimation unit 112 does not perform such position estimation, it is possible to suppress the increase in the deviation of the estimated position of worker WK over time due to drift of the motion sensor 200.
[0031] Furthermore, the position estimation unit 112 estimates that the worker WK is located at a pre-associated location, sets that location as the reference position, estimates the displacement of the worker WK using the motion sensor 200, and estimates the position of the worker WK using the reference position and the displacement. Therefore, the reference position can be set each time work is estimated, and the position of the worker WK can be estimated using the displacement from the reference position. This makes it possible to further suppress the increase in the estimated position of the worker WK over time.
[0032] Furthermore, according to the system 100 of this embodiment, the motion sensor 200 includes an inertial measuring device, which allows for the detection of large movements, including fluctuations in the skeletal structure of the worker WK.
[0033] B. Other embodiments: (B1) In the above embodiment, the work estimated by the work estimation unit 111 may be performed repeatedly at predetermined time intervals. For example, since each work performed by the factory is a routine work, the time interval for each work can be predetermined. With this configuration, since the work is performed repeatedly at predetermined time intervals, it is estimated that the worker WK is located at a predetermined location for each time interval. This makes it possible to further suppress the increase in the deviation of the estimated location of the worker WK over time. Note that the work may be of a single type or of multiple types.
[0034] (B2) In the above embodiment, the motion sensor 200 was an IMU attached to the worker WK, but the disclosure is not limited thereto. The motion sensor 200 may be a finger sensor attached to the worker WK's finger and used to detect the movement of the worker WK's finger. With this configuration, fine movements of the worker WK's finger can be detected, and the work estimation unit 111 can estimate work that includes fine movements of the finger.
[0035] (B3) In the above embodiment, the tasks performed by worker WK were described as ST1, which involves lifting the workpiece W1; ST2, which involves bolting the workpiece W1 to the workpiece; and ST3, which involves placing the workpiece W1 in the workpiece storage area SP. However, this disclosure is not limited to these tasks. Any task performed by worker WK and estimated by the work estimation unit 111 may be any task.
[0036] (B4) In the above embodiment, a part of the system 100 may be provided on an external device such as another computer. For example, one of the work estimation unit 111 and the position estimation unit 112 may be provided on another computer.
[0037] This disclosure is not limited to the embodiments described above, and can be implemented in various configurations without departing from its spirit. For example, the technical features in the embodiments corresponding to the technical features in each form described in the summary of the invention can be replaced or combined as appropriate in order to solve some or all of the above-described problems, or to achieve some or all of the above-described effects. Furthermore, if a technical feature is not described as essential in this specification, it can be deleted as appropriate. [Explanation of Symbols]
[0038] 100...System, 110...Processor, 111...Work Estimation Unit, 112...Position Estimation Unit, 120...Memory, 121...Program, 122...Work Memory Unit, 123...Position Memory Unit, 130...Input / Output Interface, 140...Internal Bus, 150...Communication Device, 160...Display Device, 200...Motion Sensor, 300...Position Estimation Device, DR...Driver, P1, P2, P3...Position, SP...Work Area, ST1, ST2, ST3...Work, W1...Workpiece, WB...Workbench, WK...Worker
Claims
1. A system for estimating the position of a worker, A motion sensor attached to the worker to detect the worker's movements, A work estimation unit that estimates the work being performed by the worker using the detected actions, A position estimation unit that estimates the position of the worker using the estimated work, the position estimation unit that estimates the worker is located at a location pre-associated with the estimated work, A system equipped with these features.
2. A system according to claim 1, wherein the operation is performed repeatedly at predetermined time intervals.
3. A system according to claim 1 or 2, The position estimation unit, After estimating that the worker is located at the aforementioned pre-associated location, the location is set as the reference location. The displacement of the worker is estimated using the motion sensor. The worker's position is estimated using the aforementioned reference position and the amount of displacement. system.
4. The system according to claim 3, The motion sensor is a system that includes an inertial measuring device attached to the worker.
5. The system according to claim 4, The aforementioned action includes the movement of the worker's fingers, The system includes a motion sensor attached to the fingers.