Remote AI Control of Autonomous Vehicles Under Incident Risk
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current autonomous vehicle systems lack effective methods to remotely monitor and control multiple independently operating vehicles to manage incident risk levels, leading to potential safety hazards due to sensor malfunctions and unforeseen situations, especially in scenarios involving pedestrians and other vulnerable road users.
Innovation Solution
Implementing a remote monitoring and control center that uses distributed sensor fusion and artificial intelligence to determine incident risk levels across multiple autonomous vehicles, taking control of vehicles when risk levels exceed safe margins and restoring safety, and returning control once safe conditions are met.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If autonomous vehicles operate independently without remote monitoring, then vehicle operational autonomy is maintained, but safety risk management capability deteriorates
Solution Approach 1:
A remote monitoring and control center is introduced as an intermediary between autonomous vehicles and human operators. The center receives sensor data from multiple vehicles, processes it through AI algorithms to assess incident risk levels, and intervenes only when necessary to take control of vehicles. This mediator enables independent vehicle operation while providing centralized safety oversight through distributed sensor fusion and AI-based risk assessment.
2Reliability
If remote monitoring of multiple autonomous vehicles is implemented, then safety oversight is improved, but system complexity increases
Solution Approach 1:
Sensor data from multiple autonomous vehicles is merged and processed together at the remote monitoring center. The system combines sensor data from various vehicles to assess incident risk levels collectively, enabling comprehensive safety oversight through data fusion rather than analyzing each vehicle in isolation. This merging approach improves safety oversight while managing system complexity through centralized processing.
Solution Approach 2:
The remote monitoring center serves multiple functions: receiving sensor data from multiple vehicles, processing data through AI algorithms, assessing incident risk levels, and controlling vehicles when necessary. This multi-functional system handles diverse tasks through a single centralized platform, improving safety oversight capability while avoiding the need for separate specialized systems for each function.
3Measurement precision
If continuous monitoring of all autonomous vehicles is performed, then incident detection capability is improved, but computational resource consumption increases
Solution Approach 1:
The system performs continuous monitoring of sensor data from all vehicles but only takes control actions when incident risk levels exceed predetermined thresholds. Rather than constantly intervening, the system monitors all data continuously and applies control only when necessary, enabling high incident detection capability while minimizing computational resource consumption through selective intervention based on AI-assessed risk levels.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Apparatus and methods relate to a controller configured to monitor incident risk levels of multiple independently governed autonomous vehicles remote from the controller simultaneously; and, in response to an unsafe incident risk level for one or more vehicles, take control of vehicles having an unsafe incident risk level, to restore a safe incident risk level; and, in response to determining incident risk has been restored to a safe level, returning control to the autonomous vehicles. Incident risk may be determined for multiple vehicles individually, and as a group, based on data from sensors distributed across multiple vehicles. Sensor data from multiple vehicles may be fused, permitting accurate incident risk determination for a vehicle group. Safety measures may be targeted by artificial intelligence to an individual vehicle or a vehicle group, to reduce incident risk by increasing separation between vehicles, or reducing vehicle speed.