Mobile Body Control in Non-Road Areas Using Pedestrian Signal Cues
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing control systems for mobile objects do not adequately consider pedestrian signals, leading to potential inability to control the mobile object in certain circumstances.
Innovation Solution
A control device for mobile objects that includes a road type recognizing unit, an acquisition unit, a traffic signal recognizing unit, and a control unit. This device recognizes the type of road, acquires auxiliary information such as alarm sounds for visually impaired persons, and uses this information to control the mobile object based on traffic signals, even when primary traffic information cannot be recognized.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If autonomous driving control is implemented using existing technologies, then the mobile body can operate autonomously, but the response speed is slow and cannot handle emergency situations appropriately
Solution Approach 1:
The system pre-trains multiple AI models (first AI model for normal driving, second AI model for emergency situations) before actual operation. This preliminary training allows the system to have pre-prepared response strategies for different scenarios, enabling fast switching during emergencies without requiring complex real-time analysis.
Solution Approach 2:
The system changes the switching threshold parameter dynamically based on situation assessment. When an emergency is detected, the threshold for switching from the first AI model to the second AI model is adjusted, allowing the system to respond appropriately to varying levels of urgency and different types of situations.
2Reliability
If multiple AI models are trained and stored for different driving situations, then emergency response capability is improved, but the device complexity increases
Solution Approach 1:
The system employs automated model selection and switching mechanisms that operate without manual intervention. The determination unit automatically assesses the current situation and selects the appropriate AI model, while the switching unit automatically transitions between models based on predefined criteria and dynamic thresholds, reducing the need for complex manual model management.
Solution Approach 2:
Multiple AI models are pre-trained and stored in the storage unit before actual operation. This preliminary preparation allows the system to have ready-to-use models for different scenarios, eliminating the need for complex real-time model training or selection algorithms during operation.
3Speed
If the switching threshold for AI models is set too low, then emergency response is faster, but normal driving interruptions increase
Solution Approach 1:
The system dynamically adjusts the switching threshold parameter based on the assessed situation. During normal driving, the threshold remains higher to prevent unnecessary switching, while during emerging emergencies, the threshold is lowered to enable faster response. This dynamic parameter adjustment optimizes both normal driving continuity and emergency response speed.
Solution Approach 2:
The switching threshold is not a fixed value but a dynamic parameter that changes based on real-time situation assessment. The determination unit continuously evaluates the current driving context and adjusts the threshold accordingly, allowing the system to adapt its switching behavior to match the actual needs of the situation.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A control device for a mobile object that is able to move in both a roadway and a predetermined area other than the roadway includes: a road type recognizing unit configured to recognize whether the mobile object is moving in the roadway or in the predetermined area on the basis of an output of a sensing device for sensing circumstances of the mobile object; an acquisition unit configured to acquire auxiliary information which is information other than traffic information indicating permission of progress and instruction of stop instructed by signaling of a pedestrian signal on which the mobile object focuses and which is indicated by the pedestrian signal and which is used to recognize the permission of progress or the instruction of stop indicated by signaling of the pedestrian signal; a traffic signal recognizing unit configured to recognize traffic information which is the permission of progress or the instruction of stop indicated by signaling of the pedestrian signal using the auxiliary information acquired by the acquisition unit; and a control unit configured to control the mobile object on the basis of the traffic information of the pedestrian signal recognized by the traffic signal recognizing unit when the mobile object is located in the predetermined area.