Autonomous Vehicle Traffic Signal Inference From Surrogate Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in navigating environments where certain aspects, such as traffic signals or obstacles, are not directly observable by their sensors, leading to potential safety issues and inefficiencies.
Innovation Solution
The integration of an inference system within the vehicle, utilizing sensor data from various sources like cameras, RADAR, and LIDAR, to infer unobservable aspects of the environment, such as traffic signal states or obstacle positions, and adjust vehicle control accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If autonomous vehicles rely solely on direct sensor observation, then device complexity is reduced, but measurement precision deteriorates when aspects of the environment are not directly observable
Solution Approach 1:
The patent introduces surrogate data as an intermediary mechanism to indirectly observe unobservable environmental aspects. Instead of directly detecting traffic signals or obscured obstacles, the system uses sensors to detect surrogate indicators (such as vehicle behavior patterns, road geometry, or environmental context) that correlate with the target aspects. This intermediary approach enables inference of hidden information without requiring direct line-of-sight sensors, thereby improving measurement precision while avoiding the complexity of omnidirectional sensing systems.
Solution Approach 2:
The patent replaces direct mechanical/optical detection systems with an inference-based computational system. Rather than using complex sensor arrays to directly observe every aspect of the environment, the system substitutes physical detection with algorithmic inference that processes surrogate data to deduce unobservable conditions. This substitution reduces device complexity by eliminating redundant sensors while maintaining or improving detection accuracy through intelligent data processing.
2Reliability
If the vehicle uses multiple sensors to observe all aspects of the environment, then measurement precision improves, but device complexity increases
Solution Approach 1:
The patent uses surrogate data as a mediator to achieve reliable environmental awareness without comprehensive direct sensing. By identifying and measuring surrogate indicators that are easier to detect and that correlate with critical safety parameters, the system achieves high reliability in navigation decisions while avoiding the complexity and cost of omnidirectional sensor coverage. For example, inferring traffic signal state from vehicle queue behavior rather than directly observing the signal.
Solution Approach 2:
The system leverages existing sensor data collected for other purposes (self-service) to infer additional environmental aspects. Rather than adding dedicated sensors for every possible observation, the patent reuses data from sensors already present for primary functions (e.g., using LIDAR data for both obstacle detection and traffic signal inference), thereby improving reliability without increasing device complexity.
3Measurement precision
If the vehicle waits for direct observation of traffic signals, then measurement precision is ensured, but loss of time increases
Solution Approach 1:
The patent applies preliminary action by inferring traffic signal state in advance based on surrogate data before the vehicle reaches the intersection. By analyzing patterns in surrogate indicators (such as the behavior of preceding vehicles, road geometry, or timing patterns) as the vehicle approaches, the system predicts the traffic signal state ahead of time, allowing smooth deceleration and elimination of unnecessary stops. This maintains measurement precision through continuous inference while eliminating time loss from reactive stopping.
Solution Approach 2:
The system uses feedback from surrogate data to continuously update its inference of traffic signal state as the vehicle approaches the intersection. By monitoring changes in surrogate indicators over time and comparing them against expected patterns, the system refines its prediction of the signal state, maintaining high accuracy while optimizing the timing of deceleration to minimize stopping time.
Data Source
AI summary
A vehicle configured to operate in an autonomous mode can obtain sensor data from one or more sensors observing one or more aspects of an environment of the vehicle. At least one aspect of the environment of the vehicle that is not observed by the one or more sensors could be inferred based on the sensor data. The vehicle could be controlled in the autonomous mode based on the at least one inferred aspect of the environment of the vehicle.