Remote Vehicle Takeover Using Compressed Context Playback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous vehicles require manual assistance in unpredictable situations, such as road construction or severe weather, which can be inefficient and costly, and operators need to be prepared physically and psychologically for remote control, but current systems do not effectively prepare operators before taking over.
Innovation Solution
A method and system that detects the need for manual assistance, adjusts vehicle data playback to prepare the operator, and adapts vehicle operation to ensure the operator is ready at the takeover position, using vehicle data such as video streams, audio, and haptic data, and considers operator availability and state to enable efficient remote control.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the operator is provided with historical vehicle data for preparation before remote control, then the operator's readiness is improved, but the time required for operator preparation increases
Solution Approach 1:
The system performs preliminary actions by providing the operator with historical vehicle data (video, audio, haptic feedback) before the actual remote control operation begins. This allows the operator to familiarize themselves with the vehicle environment and potential situations in advance, improving readiness without delaying the actual control task
Solution Approach 2:
The system dynamically adapts the preparation process by adjusting the amount and type of historical data provided based on the specific situation, operator performance, and time constraints. The playback speed and data selection are modified in real-time to optimize both preparation quality and time efficiency
2Productivity
If the vehicle operates autonomously without manual assistance, then productivity is improved, but the ability to handle unpredictable situations deteriorates
Solution Approach 1:
The system introduces a remote operator as an intermediary between the autonomous vehicle system and unpredictable situations. The vehicle operates autonomously under normal conditions, but can summon remote human assistance when encountering situations beyond its autonomous capabilities, combining the efficiency of automation with the adaptability of human judgment
Solution Approach 2:
The system implements feedback mechanisms where the autonomous vehicle monitors its own performance and environmental conditions, automatically detecting when manual assistance is needed. The vehicle provides continuous status feedback to the remote operator, enabling the operator to understand the situation and provide appropriate guidance or take control when necessary
Data Source
Figure 1~2B
Figure 3~4
Figure 5A~5B
AI summary
It is provided a method for enabling remote control of a vehicle with autonomous propulsion capability. The method is performed by a vehicle data provider and comprises: detecting a need for manual assistance of the vehicle by an operator being remote from the vehicle; obtaining a stream of vehicle data, the vehicle data relating to a time prior to when remote control starts; modifying the vehicle data, which comprises adjusting a duration of playback of the vehicle data; providing the modified vehicle data for playback to the operator; providing, once the playback of modified vehicle data has ended, vehicle data in real-time to the operator; and enabling remote control of the vehicle by the operator.