Robot Motion Planning Hardware With Parallel PRM Collision Circuits
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing motion planning algorithms for robots are computationally intensive and power-hungry, making them insufficient for real-time operation in environments with obstacles, particularly in applications like healthcare, personal assistance, and search-and-rescue where fast and efficient navigation is crucial.
Innovation Solution
A specialized hardware architecture that uses a probabilistic roadmap (PRM) and collision detection units with Boolean logic gates and lookup tables to enable real-time motion planning by processing environmental data from sensors, identifying collision-free paths, and optimizing the hardware design for specific robots and scenarios to reduce power consumption and cost.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If motion planning algorithms are implemented using general-purpose processors or GPUs, then computational capability is sufficient, but power consumption becomes too high for many target applications
Solution Approach 1:
The motion planning system is segmented into two distinct parts: an offline phase that pre-computes a probabilistic roadmap (PRM) and stores it in memory, and an online phase that performs real-time query processing. This segmentation allows the computationally intensive collision detection to be performed once offline, while real-time operation uses pre-computed data, dramatically reducing power consumption during actual robot operation.
Solution Approach 2:
The system performs preliminary action by pre-computing the probabilistic roadmap and collision detection results before real-time operation begins. The offline phase generates a comprehensive PRM that captures the robot's configuration space and collision-free paths, which is then stored for rapid retrieval during real-time queries, eliminating the need for repeated computationally intensive calculations.
2Speed
If motion planning algorithms are implemented in software on general-purpose processors, then flexibility is maintained, but real-time performance is insufficient
Solution Approach 1:
The system employs dynamic reconfiguration capabilities through FPGAs, allowing the hardware architecture to be adapted and optimized for specific robot types and motion planning requirements. This dynamic aspect enables the system to achieve real-time performance by configuring specialized collision detection circuits tailored to the specific robot's kinematics and environment, while maintaining flexibility for different applications.
Solution Approach 2:
The system replaces general-purpose software-based motion planning with specialized hardware circuits designed specifically for collision detection and PRM query processing. This substitution of mechanical/computational systems with dedicated hardware accelerates the real-time performance by performing collision checks in parallel using purpose-built logic circuits rather than sequential software execution.
3Use of energy by moving object
If specialized hardware is designed for specific robots, then power consumption and cost are reduced, but adaptability to different robots decreases
Solution Approach 1:
The system achieves universality through a hierarchical architecture where the offline PRM generation and online query processing framework remains general and applicable to any robot, while only the specific collision detection circuits need to be customized. The FPGAs can be reconfigured for different robot types, and the overall system framework handles various robot kinematics, degrees of freedom, and environment types, making the solution both energy-efficient and adaptable.
Data Source
Figure 1A~1B
Figure 1C~2
Figure 3
AI summary
Specialized robot motion planning hardware and methods of making and using same are provided. A robot-specific hardware can be designed using a tool that receives a robot description comprising a collision geometry of a robot, degrees of freedom for each joint of the robot, and joint limits of the robot; receives a scenario description; generates a probabilistic roadmap (PRM) using the robot description and the scenario description; and for each edge of the PRM, produces a collision detection unit comprising a circuit indicating all parts of obstacles that collide with that edge. The hardware is implemented as parallel collision detection units that provide collision detection results used to remove edges from the PRM that is searched to find a path to a goal position.