Autonomous Vehicle Sensor Unavailability Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous driving systems face safety and comfort issues due to sensor unavailability and communication disruptions, leading to unplanned stops and stressful transitions to manual control.
Innovation Solution
A vehicle system that predicts sensor unavailability based on performance thresholds and environmental conditions, activating countermeasures such as route alternation, dead reckoning, or stopping to ensure safety and comfort by alerting the driver and compensating with other sensors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the autonomous vehicle stops when a sensor becomes unavailable, then safety is improved, but driver comfort deteriorates due to unplanned stops and stressful transitions to manual control
Solution Approach 1:
The system performs preliminary actions by predicting sensor unavailability before it actually occurs. The processing unit calculates the probability and timing of sensor failure using current sensor data and environmental information, then activates countermeasures in advance. This allows the vehicle to prepare for sensor failure gracefully, maintaining driver comfort while ensuring safety through proactive rather than reactive responses.
Solution Approach 2:
The system dynamically adjusts its response based on the predicted probability and timing of sensor unavailability. Rather than a fixed stop-or-continue decision, the system selects from multiple countermeasures (generating alternative routes, activating dead reckoning, stopping) based on real-time assessments. This dynamic approach optimizes the balance between safety and driver comfort according to specific situation parameters.
2Ease of operation
If the vehicle activates countermeasures when sensor unavailability is predicted, then driver comfort is improved by avoiding unplanned stops, but system complexity increases
Solution Approach 1:
The system segments the sensor unavailability prediction into distinct components: probability calculation, timing estimation, and countermeasure selection. The processing unit handles each aspect separately through dedicated algorithms, making the complex overall function more manageable and maintainable. This modular approach to handling sensor failure prediction reduces system complexity compared to a monolithic solution.
Solution Approach 2:
The system introduces an intermediary prediction layer between sensor data and control actions. Rather than directly responding to sensor failures, the processing unit acts as an intermediary that predicts future sensor availability and mediates the selection of appropriate countermeasures. This intermediary layer simplifies the overall system architecture by decoupling sensor monitoring from control decisions.
3Measurement precision
If the vehicle relies on real-time sensor data for autonomous driving, then navigation accuracy is improved, but reliability deteriorates when sensors become unavailable or communication is interrupted
Solution Approach 1:
The system performs preliminary assessment of sensor reliability by calculating the probability of sensor unavailability before actual failures occur. Using current sensor performance data and environmental conditions, the system predicts future sensor availability and prepares alternative navigation strategies in advance. This preliminary action ensures that navigation accuracy is maintained by having backup methods ready before sensors fail.
Solution Approach 2:
The system changes operational parameters based on predicted sensor availability. When sensor unavailability is predicted, the system transitions from relying on real-time sensor data to alternative methods such as dead reckoning or pre-planned routes. This parameter change in the navigation approach maintains reliability by adapting to anticipated sensor conditions while preserving navigation accuracy through appropriate method selection.
Data Source
Figure 1a~1b
Figure 2a~2b
Figure 3
AI summary
A vehicle (1), a vehicle system (2) and a method (100) for increasing at least one of safety and comfort during autonomous driving is provided. The vehicle system (2) comprises an autonomous drive arrangement (3) with a plurality of sensors (5), a vehicle control arrangement (8) and a positioning system (4).The vehicle system (2) is arranged to determine an estimated probability that at least one sensor (5) will become unavailable, or an estimated time/ distance ahead until at least one sensor (5) is determined to become unavailable. The vehicle system (2) is further arranged to activate at least one countermeasure based on at least one of the estimated probability, the estimated time and the estimated distance.