Remote Micro-Authorization for Autonomous Vehicle Phantom Stops
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles often incorrectly cease forward advancement due to erroneous sensor system outputs or flawed control algorithms, leading to phantom object detection, which can result in undesirable experiences for passengers and hinder the vehicle's primary purpose of navigating to a destination without human control.
Innovation Solution
A computing system that generates instructions to incrementally advance the autonomous vehicle by deactivating constraints inhibiting its propulsion, allowing it to move a predefined distance with remote assistance, using a remote computing system to receive signals from the vehicle, identify constraints, and transmit return signals with instructions to deactivate them, ensuring safe navigation and minimizing manual intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the autonomous vehicle executes default instructions to maintain mechanical systems in an inactive state until an affirmative sensor signal is received, then safety is improved, but the vehicle cannot advance beyond its current location when phantom objects are detected
Solution Approach 1:
A remote computing system acts as an intermediary between the autonomous vehicle and human operators. When the vehicle detects a phantom object and cannot safely advance, it transmits sensor data and constraint information to the remote computing system, which processes the information and provides guidance instructions to the vehicle, enabling safe resolution without requiring full manual control
Solution Approach 2:
The control system is segmented into autonomous operation mode and remote assistance mode. The autonomous vehicle operates independently under normal conditions, but when constraints are activated due to phantom object detection, it transitions to remote assistance mode where a remote computing system provides segmented guidance instructions for specific actions rather than full manual takeover
2Productivity
If manual control is allowed to advance the autonomous vehicle beyond its current location, then vehicle productivity is improved, but the primary purpose of autonomous navigation is compromised
Solution Approach 1:
Instead of full manual control, the system implements partial human intervention through remote assistance. The remote computing system provides specific guidance instructions for resolving phantom object detections, allowing the vehicle to advance past obstacles without requiring continuous manual control, thus maintaining most autonomous functionality while enabling productivity
3Reliability
If the autonomous vehicle ceases forward advancement due to phantom object detection, then false positive safety concerns are addressed, but passenger experience deteriorates
Solution Approach 1:
The system implements feedback loops where sensor data is continuously monitored, phantom objects are identified and verified, and the vehicle's response is communicated to passengers. When phantom objects are detected, the remote computing system processes the situation and provides feedback instructions, allowing passengers to understand the situation and the vehicle's intended actions, thereby maintaining trust and comfort
Data Source
AI summary
Micro-authorization of remote assistance for an autonomous vehicle is described herein. A constraint that inhibits propulsion by a mechanical system of the autonomous vehicle is activated by a computing system of the autonomous vehicle, wherein a signal that identifies the activated constraint is transmitted from the autonomous vehicle to a remote computing system. The remote computing system generates instructions to deactivate the activated constraint. A return signal is transmitted from the remote computing system that specifies instructions to deactivate the constraint and a distance to desirably advance the autonomous vehicle. The computing system of the autonomous vehicle deactivates the constraint and the mechanical system is controlled to advance the autonomous vehicle when signal latency is less than a predetermined threshold duration of time.