Autonomous Vehicle Remote Support Validation Under Communication Delay
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Delays in information transmission and reception between a vehicle and a remote support operator in autonomous driving systems can lead to situations where the vehicle receives ineffective support, as the environment around the vehicle changes significantly during the delay, causing the remote support to be no longer applicable.
Innovation Solution
A travel management apparatus that acquires sensor information, generates prediction information for peripheral objects, and transmits remote support requests, receiving instructions only if the predicted and actual positions of these objects correspond, ensuring the support remains effective by confirming the correspondence of traffic states before implementing the remote support.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If remote support is transmitted from remote support operator to vehicle via communication network, then remote support capability is achieved, but communication delay occurs causing the support to become ineffective
Solution Approach 1:
The system performs preliminary actions by continuously acquiring sensor information and generating prediction information about peripheral objects before the remote support is transmitted. This allows the vehicle to have updated environmental data ready when the remote support arrives, reducing the effective impact of communication delay.
Solution Approach 2:
The system implements feedback by comparing prediction information (generated before remote support transmission) with actual sensor information (acquired when remote support is received). This feedback mechanism determines whether the remote support remains valid or needs to be rejected, ensuring safety despite communication delays.
2Reliability
If remote support request is transmitted to remote support operator, then determination support is obtained, but delay occurs making the situation different from what remote support operator recognized
Solution Approach 1:
The vehicle performs preliminary actions by continuously monitoring the environment and generating prediction information during the delay period. This ensures that when remote support is received, the vehicle can verify whether the support is still valid for the current situation.
Solution Approach 2:
A feedback mechanism compares the predicted situation (from prediction information) with the actual current situation (from sensor information). This feedback determines whether to accept or reject the remote support, maintaining determination accuracy despite delays.
3Reliability
If the vehicle waits for remote support before making travel decisions, then human oversight is achieved, but travel efficiency decreases due to delay
Solution Approach 1:
The system performs preliminary actions by continuously acquiring sensor information and generating prediction information in advance. This allows the vehicle to maintain awareness of its environment without waiting for remote support, improving travel efficiency while still enabling human oversight when needed.
Solution Approach 2:
The feedback mechanism allows the vehicle to automatically determine whether remote support is still valid upon receipt. This reduces the need to wait passively for remote support decisions, improving travel efficiency while maintaining reliable human oversight when the situation has not changed significantly.
Data Source
AI summary
The present disclosure provides to an apparatus for managing travel of a vehicle having an autonomous travel function. The apparatus executes the following steps. The apparatus sequentially acquires sensor information from a recognition sensor mounted on the vehicle. The apparatus sequentially generates prediction information showing a position of an object at a future time point based on the sensor information acquired sequentially. The apparatus transmits a remote support request to a remote support operator. The apparatus receives a remote support given in response to the remote support request by the remote support operator. The apparatus makes the vehicle autonomously travel in accordance with the remote support in response to confirming that an actual position of an object obtained from sensor information and a predicted position of an object obtained from prediction information correspond.