Probability-Based Obstacle Avoidance for Dynamic Hazard Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional obstacle avoidance systems face challenges in integrating dynamic obstacle detection with stationary obstacle databases, leading to increased computational resource demands and inefficiencies in autonomous vehicles.
Innovation Solution
A probability-based obstacle database system that defines obstacles in terms of spatial and temporal probability distributions, allowing efficient comparison with on-board sensor data to reduce processing load and enhance safety by focusing on potential hazards.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional obstacle databases focused on stationary objects are used, then avoidance of stationary obstacles is enabled, but detection and avoidance of non-stationary obstacles requires additional components increasing system complexity
Solution Approach 1:
The patent transforms the traditional stationary obstacle database into a universal database that handles both stationary and non-stationary obstacles through probability distribution functions. The database structure is modified to accommodate dynamic objects by incorporating temporal components and probability-based representations, allowing a single database system to perform multiple obstacle types detection without requiring separate dedicated systems for each obstacle category.
Solution Approach 2:
The patent changes the fundamental parameters of obstacle representation from deterministic coordinates to probability distribution functions with spatial and temporal components. This parameter transformation allows the system to represent both stationary objects (with high confidence, fixed location distributions) and non-stationary objects (with lower confidence, moving location distributions) using the same mathematical framework, thereby unifying the database structure.
2Reliability
If additional components are added for dynamic obstacle detection, then non-stationary obstacle detection capability is improved, but computational resource demands increase
Solution Approach 1:
The patent merges the functions of separate stationary and dynamic obstacle detection systems into a unified probability-based database framework. By combining spatial and temporal probability distributions into a single computational model, the system eliminates redundant processing that would occur if separate detection systems were used, thereby reducing overall computational resource consumption while maintaining comprehensive obstacle detection capability.
3Measurement precision
If comprehensive sensor data processing is implemented for dynamic obstacles, then detection precision is improved, but processing time and computational load increase
Solution Approach 1:
The patent implements preliminary action by pre-computing and storing probability distribution functions for obstacles in the database before real-time operation. Spatial and temporal probability models are established in advance for known obstacle zones, allowing the system to quickly compare sensor data against pre-established probability thresholds rather than performing complex real-time calculations, thereby maintaining high detection precision while minimizing processing time.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A system includes an interface configured to receive sensor data corresponding to a potential hazard associated with a travel path of a vehicle. The system also includes an obstacle database including probability distributions of potential obstacles. The system further includes one or more processors coupled to the interface and configured to generate a probability index associated with the potential hazard and to determine whether to modify the travel path based on the probability index and the probability distributions.