Optimized directed roadmap graph for multi-agent path finding using stochastic gradient descent

@article{Henkel2020OptimizedDR,
  title={Optimized directed roadmap graph for multi-agent path finding using stochastic gradient descent},
  author={Christian Henkel and Marc Toussaint},
  journal={Proceedings of the 35th Annual ACM Symposium on Applied Computing},
  year={2020}
}
We present a novel approach called Optimized Directed Roadmap Graph (ODRM). It is a method to build a directed roadmap graph that allows for collision avoidance in multi-robot navigation. This is a highly relevant problem, for example for industrial autonomous guided vehicles. The core idea of ODRM is, that a directed roadmap can encode inherent properties of the environment which are useful when agents have to avoid each other in that same environment. Like Probabilistic Roadmaps (PRMs), ODRM… 

Figures from this paper

CTRMs: Learning to Construct Cooperative Timed Roadmaps for Multi-agent Path Planning in Continuous Spaces

CTRMs enable each agent to focus on its important locations around potential solution paths in a way that considers the behavior of other agents to avoid inter-agent collisions, while being augmented in the time direction to make it easy to derive a “timed” solution path.

Avoidance Critical Probabilistic Roadmaps for Motion Planning in Dynamic Environments

This work proposes a self-supervised methodology for learning to identify regions frequently used for obstacle avoidance from local environment features and leverages a neural network to generate hierarchical probabilistic roadmaps termed Avoidance Critical Probabilistic Roadmaps (ACPRM).

Automated Roadmap Graph Creation and MAPF Benchmarking for Large AGV Fleets

  • Jonas StenzelLea Schmitz
  • Computer Science
    2022 8th International Conference on Automation, Robotics and Applications (ICARA)
  • 2022
A concept and first evaluation of an automated topology creation and two different Multi-Agent Path Finding approaches for large robot fleets with the focus on logistic applications are presented.

Reinforcement-Learning-Based Route Generation for Heavy-Traffic Autonomous Mobile Robot Systems

The paper proposes a reinforcement learning approach where an agent builds the routes on a given layout while being rewarded according to different criteria based on the desired characteristics of the system to improve the operation and cooperation of multiple robots in their shared environment.

Multilevel Motion Planning: A Fiber Bundle Formulation

The terminology of fiber bundles is introduced, in particular the notion of restrictions and sections, which is used to develop novel multilevel motion planning algorithms, which are called QRRT and QMP and shown to be probabilistically complete and almost-surely asymptotically optimal.

References

SHOWING 1-10 OF 42 REFERENCES

Probabilistic roadmaps for path planning in high-dimensional configuration spaces

Experimental results show that path planning can be done in a fraction of a second on a contemporary workstation (/spl ap/150 MIPS), after learning for relatively short periods of time (a few dozen seconds).

Probabilistic Roadmaps for Path Planning in High-Dimensional Configuration Spaces

Experimental results show that the proposed roadmap method can eeciently solve problems which are beyond the capabilities of other existing methods.

Distributed Roadmaps for Robot Navigation in Sensor Networks

A distributed sampling-based planning algorithm is proposed, where every sensor node creates a local roadmap in its locally sensed environment; these local roadmaps are “stitched” together by passing messages among nodes and forming a larger implicit roadmap without having a global representation.

Optimal Multirobot Path Planning on Graphs: Complete Algorithms and Effective Heuristics

The combination of ILP model based algorithms and the heuristics proves to be highly effective, allowing the computation of 1.x-optimal solutions for problems containing hundreds of robots, densely populated in the environment, often in just seconds.

Multi-agent RRT: sampling-based cooperative pathfinding

This work proposes MA-RRT*, a novel algorithm for multi-agent path planning that builds upon a recently proposed asymptotically-optimal sampling-based algorithm for finding single-agent shortest path called RRT*.

Finding a needle in an exponential haystack: Discrete RRT for exploration of implicit roadmaps in multi-robot motion planning

The pathfinding algorithm, discrete-RRT (dRRT), is an adaptation of the celebrated RRT algorithm for the discrete case of a graph, and it enables a rapid exploration of the high-dimensional configuration space by carefully walking through an implicit representation of the tensor product of roadmaps for the individual robots.

Optimal Multi-Robot Path Planning on Graphs: Structure and Computational Complexity

This paper designs complete algorithms and efficient heuristics for optimizing all four objectives of MPP, capable of solving MPP optimally or near-optimally for hundreds of robots in challenging setups.

Multi-agent Generalized Probabilistic RoadMaps: MAGPRM

The algorithm is used to solve multi-agent motion planning problems involving 2-dimensional and 3-dimensional agents in stochastic maps with uncertainty in the motion model and indicates that the algorithm successfully solves the problem under uncertainty, and generates a solution having high probability of success.

Motion planning using dynamic roadmaps

  • Marcelo KallmannM. Matarić
  • Computer Science
    IEEE International Conference on Robotics and Automation, 2004. Proceedings. ICRA '04. 2004
  • 2004
This paper focuses on analyzing the tradeoffs between maintaining dynamic roadmaps and applying an on-line bidirectional rapidly-exploring random tree (RRT) planner alone, which requires no preprocessing or maintenance.

Real-time navigation of independent agents using adaptive roadmaps

A novel algorithm for navigating a large number of independent agents in complex and dynamic environments by taking into account dynamic obstacles and inter-agents interaction forces to continuously update the roadmap by using a physically-based agent dynamics simulator.