Control device
The control device resolves AI model conflicts by prioritizing AI models using a selection and comparison mechanism, ensuring autonomous moving bodies navigate effectively without computational overload.
Patent Information
- Application Number
- JP2024041153
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2024-03-15
- Publication Date
- 2025-09-29
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Conflicts between multiple AI models can lead to improper functioning when no pre-determined rule exists for prioritization, causing autonomous moving bodies to become stuck in a loop of stopping and starting due to inability to establish superiority or inferiority.
A control device with a selection and comparison mechanism that determines priority between AI models based on a predetermined rule, selecting and comparing values when moving bodies approach within a certain distance to resolve conflicts and prevent collisions.
Resolves conflicts between AI models efficiently with minimal computational load, allowing autonomous moving bodies to navigate without getting stuck, by establishing priority through simple logical operations.
Smart Images
Figure 2025141284000001_ABST
Abstract
Description
[Technical Field]
[0001] The present invention relates to the technical field of control devices. [Background technology]
[0002] As a technology related to this type of device, for example, a technology for providing services using an AI (Artificial Intelligence) model has been proposed (see Patent Document 1). [Prior art documents] [Patent documents]
[0003] [Patent Document 1] International Publication No. 2021 / 084810 Summary of the Invention [Problem to be solved by the invention]
[0004] When multiple AI models exist, there is a possibility that a conflict may arise between one AI model and another. For example, if a rule is determined in advance that determines whether one AI model or another AI model should be applied in the event of a conflict, then the rule can be followed. On the other hand, if no equivalent rule is determined, there is a technical problem in that the function using the AI model may not be properly performed.
[0005] The present invention has been made in consideration of the above problems, and an object of the present invention is to provide a control device that can resolve conflicts between AI models. [Means for solving the problem]
[0006] A control device according to one embodiment of the present invention is a control device that causes a moving body to move autonomously using an AI model, and is equipped with a selection means that selects one value from a plurality of predetermined values when the distance between the moving body that is moving autonomously and another moving body that is moving autonomously using an AI model is less than a predetermined distance, and a comparison means that compares the selected value with a value selected by the other moving body from the plurality of values. [Brief explanation of the drawings]
[0007] [Figure 1] FIG. 2 is a block diagram showing an example of the configuration of a control device according to the embodiment. [Figure 2] 4 is a flowchart illustrating an example of an operation of the control device according to the embodiment. DETAILED DESCRIPTION OF THE INVENTION
[0008] An embodiment of the control device will be described with reference to Fig. 1 and Fig. 2. In Fig. 1, the control device 10 is mounted on a moving body 1. The control device 20 is mounted on a moving body 2. The control device 10 may use an AI model to autonomously move the moving body 1. The control device 20 may use an AI model to autonomously move the moving body 2. The moving bodies 1 and 2 may be, for example, self-driving vehicles, AGVs (Automatic Guided Vehicles), robots, etc. The moving bodies 1 and 2 may be equipped with various sensors, for example, cameras, radars, etc.
[0009] An example of the autonomous movement of the mobile body 1 will be described. For example, if an obstacle is present ahead of the mobile body 1, the mobile body 1 may move so as to avoid a collision with the obstacle. The obstacle may be an object whose position does not change, or may be a moving body (e.g., a person, a vehicle, a robot, etc.). In this case, the control device 10 may use an AI model to set a travel path for the mobile body 1 that allows the mobile body 1 to avoid a collision with the obstacle. If the obstacle is a moving body, the control device 10 may set a travel path for the mobile body 1, taking into account a range in which the moving body as an obstacle may be present after a predetermined time. The control device 10 may control the mobile body 1 so that the mobile body 1 moves along the set travel path. If it is difficult to avoid a collision between the mobile body 1 and the obstacle, the mobile body 1 may stop. The mobile body 2 may also move autonomously, similar to the mobile body 1. For this reason, a description of an example of the autonomous movement of the mobile body 2 will be omitted.
[0010] For example, as shown in Fig. 1, when moving body 1 and moving body 2 are approaching each other in an opposing manner, it may not be possible to set a travel route that can avoid a collision between moving body 1 and moving body 2 due to the fact that moving bodies 1 and 2 are moving. As a result, moving bodies 1 and 2 may stop.
[0011] As described above, the travel route may be set taking into consideration the range in which a moving body as an obstacle may be located after a predetermined time. For example, if the range in which moving body 2 may be located after a predetermined time narrows due to moving body 2 stopping, control device 10 may set a travel route for moving body 1 that allows moving body 1 to avoid collision with moving body 2. Similarly, if the range in which moving body 1 may be located after a predetermined time narrows due to moving body 1 stopping, control device 20 may set a travel route for moving body 2 that allows moving body 2 to avoid collision with moving body 1.
[0012] When moving body 1 starts moving along a set travel route and moving body 2 starts moving along the set travel route, the range in which moving body 2 may exist as an obstacle to moving body 1 after a predetermined time period expands, and the range in which moving body 1 may exist as an obstacle to moving body 2 after a predetermined time period expands. As a result, moving bodies 1 and 2 may stop again. For example, if this happens repeatedly, moving bodies 1 and 2 may become unable to move. This is because no superiority or inferiority is established between moving body 1 and moving body 2 (in other words, no superiority or inferiority is established between the AI model used by control device 10 and the AI model used by control device 20).
[0013] In this embodiment, a predetermined rule is used to determine which of the AI models used by the control device 10 and the AI model used by the control device 20 is to be prioritized. The predetermined rule may be, for example, "1 is stronger than 2," "2 is stronger than 3," and "3 is stronger than 1."
[0014] As shown in FIG. 1 , the control device 10 includes a selection unit 11 and a comparison unit 12. The control device 20 includes a selection unit 21 and a comparison unit 22. When the distance between the moving object 1 and the moving object 2 becomes equal to or shorter than a predetermined distance, the selection unit 11 of the control device 10 selects one value from a plurality of values (here, "1", "2", and "3"). Similarly, the selection unit 21 of the control device 20 selects one value from the plurality of values. The control device 10 may transmit the one value selected by the selection unit 11 to the control device 20. Similarly, the control device 20 may transmit the one value selected by the selection unit 21 to the control device 20. As a result, the control device 10 may receive the one value selected by the selection unit 21 of the control device 20. Similarly, the control device 20 may receive the one value selected by the selection unit 11 of the control device 10.
[0015] The comparison unit 12 of the control device 10 may compare the one value selected by the selection unit 11 with the one value selected by the selection unit 21 of the control device 20 according to the above-mentioned predetermined rule. The comparison unit 22 of the control device 10 may compare the one value selected by the selection unit 21 with the one value selected by the selection unit 11 of the control device 10 according to the predetermined rule. For example, the one value selected by the selection unit 11 may be "2" and the one value selected by the selection unit 21 may be "3." In this case, the comparison unit 12 of the control device 10 may determine that the control device 10 has won. On the other hand, the comparison unit 22 of the control device 20 may determine that the control device 20 has lost. In this case, the control device 10 may determine that the control device 10 has priority over the control device 20. Similarly, the control device 20 may determine that the control device 10 has priority over the control device 20. In other words, it may be determined that the AI model used by the control device 10 has priority over the AI model used by the control device 20.
[0016] For example, moving body 2 may stop until moving body 1 passes around moving body 2. Alternatively, moving body 2 may move to give way to moving body 1. In this case, moving body 2 may move in the width direction of the passage and / or move backward. As a result, for example, it is possible to prevent both moving bodies 1 and 2 from becoming unable to move.
[0017] The operation of the control device 10 will be described with reference to the flowchart of Fig. 2. Note that the operation of the control device 20 may be similar to that of the control device 10, and therefore a description of the operation of the control device 20 will be omitted.
[0018] 2, the control device 10 determines whether the distance between the moving body 1 and the moving body 2 is equal to or less than a predetermined distance (step S101). The distance between the moving body 1 and the moving body 2 may be estimated based on the output of a sensor such as a radar. If it is determined in the processing of step S101 that the distance between the moving body 1 and the moving body 2 is not equal to or less than the predetermined distance (step S101: No), the operation shown in FIG. 2 may be ended.
[0019] In the process of step S101, if it is determined that the distance between the moving object 1 and the moving object 2 is equal to or shorter than a predetermined distance (step S101: Yes), the selector 11 of the control device 10 selects one value from a plurality of values (step S102). The control device 10 may receive the one value selected by the selector 21 of the control device 20. Thereafter, the comparator 12 of the control device 10 compares the one value selected by the selector 11 with the one value selected by the selector 21 of the control device 20 in accordance with the above-mentioned predetermined rule (step S103). The control device 10 may determine which of the control devices 10 and 20 is to be prioritized based on the comparison result of step S103.
[0020] (Technical Effects) In this embodiment, a value selected from a plurality of values is compared to determine which of the AI models used by the control device 10 and the control device 20 is to be prioritized. Therefore, this embodiment can resolve conflicts between AI models. Additionally, selecting a value from a plurality of values and comparing the selected values can be achieved with relatively simple logic. This can prevent an increase in the computational load on the control devices 10 and 20, and can output results relatively quickly.
[0021] In the above-described embodiment, an AI model used for the autonomous movement of a moving object is described, but the present invention can also be applied to, for example, resolving conflicts between AI models used for game control, and resolving conflicts between AI models used for control in VR (Virtual Reality) and the Metaverse.
[0022] Aspects of the invention derived from the above-described embodiments will be described below.
[0023] A control device according to one aspect of the invention is a control device for autonomously moving a moving body using an AI model, and includes: a selection means for selecting one value from a plurality of predetermined values when the distance between the autonomously moving moving body and another moving body that is also autonomously moving using the AI model is equal to or shorter than a predetermined distance; and a comparison means for comparing the selected value with a value selected by the other moving body from the plurality of values. In the above embodiment, the "selection unit 11" and the "selection unit 21" correspond to an example of the "selection means," and the "comparison unit 12" and the "comparison unit 22" correspond to an example of the "comparison means."
[0024] The present invention is not limited to the above-described embodiments, but can be modified as appropriate within the scope of the claims and the gist or idea of the invention as can be read from the entire specification, and control devices involving such modifications are also included in the technical scope of the present invention. [Explanation of symbols]
[0025] 1, 2... Mobile body, 10, 20... Control device, 11, 21... Selection unit, 12, 22... Comparison unit
Claims
[Claim 1] A control device for autonomously moving a moving object using an AI model, a selection means for selecting one value from a plurality of predetermined values when a distance between the one autonomously moving moving body and another autonomously moving body using an AI model is equal to or shorter than a predetermined distance; a comparison means for comparing the selected one value with a value selected by the other moving body from the plurality of values; A control device comprising:
Citation Information
Patent Citations
Information processing device, information processing method, and artificial intelligence model manufacturing method
WO2021084810A1