Robot Collision Detection Using Persistent and Transient Obstacles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing motion planning systems face challenges in performing collision detection efficiently, especially in dynamic environments with changing obstacles, while also requiring low cost, low energy consumption, and limited storage.
Innovation Solution
The implementation of collision detection hardware, software, and/or firmware that determines whether a robot pose or motion causes collisions with itself or obstacles, using data structure representations of the robot and environment, including persistent and transient obstacles, to facilitate efficient collision detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional collision detection algorithms are used, then collision detection can be performed, but the system requires high computational resources, high energy consumption, and significant storage capacity
Solution Approach 1:
The patent segments the environment into persistent obstacles (known at configuration time) and transient obstacles (detected at runtime), allowing the system to use different data structures and algorithms for each type. This segmentation enables efficient collision detection by preprocessing persistent obstacles into compact representations while handling transient obstacles with lighter-weight methods, thereby reducing overall computational and energy requirements while maintaining detection accuracy.
2Measurement precision
If traditional collision detection algorithms are used, then collision detection can be performed, but the system has high device complexity and requires expensive hardware
Solution Approach 1:
The patent performs preliminary actions by pre-processing persistent obstacles at configuration time to create compact data structures (such as convex decompositions, bounding volumes, or occupancy grids). This preprocessing reduces the computational burden during runtime collision detection, allowing the system to use simpler, less expensive hardware while maintaining high detection accuracy. The complex computations are moved to configuration time when the robot is not operational.
3Measurement precision
If traditional collision detection algorithms are used, then collision detection can be performed, but the system has high latency and cannot operate in real-time
Solution Approach 1:
The patent implements a dynamic collision detection system that adapts its approach based on the robot's motion state. For slow or static scenarios, it uses more computationally intensive but accurate methods. For fast-moving scenarios, it switches to simplified algorithms with lower latency. The system also dynamically updates only the portions of the environment that have changed, rather than re-evaluating the entire scene, thereby reducing detection latency while maintaining accuracy.
4Measurement precision
If traditional collision detection algorithms are used, then collision detection can be performed, but the system requires large storage capacity for environment data
Solution Approach 1:
The patent creates simplified copies or representations of persistent obstacles at configuration time, storing compact data structures such as convex decompositions, bounding volume hierarchies, or occupancy grids instead of detailed geometric models. These simplified representations retain sufficient information for accurate collision detection while occupying minimal storage space. The system only stores essential geometric features needed for collision queries, discarding unnecessary detail.
Data Source
AI summary
Collision detection useful in motion planning for robotics advantageously employs data structure representations of robots, persistent obstacles and transient obstacles in an environment in which a robot will operate. Data structures may take the form of hierarchical data structures ((e.g., octrees, sets of volumes or boxes (e.g., a tree of axis-aligned bounding boxes (AABBs), a tree of oriented (not axis-aligned) bounding boxes, or a tree of spheres)) or non-hierarchical data structures (e.g., Euclidean Distance Fields) Such can result in computational efficiency, reduce memory requirements, and lower power consumption. The collision detection can take the form as a standalone function, providing a Boolean result that can be employed in executing any of a variety of different motion planning algorithms.