Work Smarter Not Harder: Simple Imitation Learning with CS-PIBT Outperforms Large Scale Imitation Learning for MAPF
September 22, 2024 Β· Declared Dead Β· π IEEE International Conference on Robotics and Automation
"No code URL or promise found in abstract"
Evidence collected by the PWNC Scanner
Authors
Rishi Veerapaneni, Arthur Jakobsson, Kevin Ren, Samuel Kim, Jiaoyang Li, Maxim Likhachev
arXiv ID
2409.14491
Category
cs.MA: Multiagent Systems
Cross-listed
cs.RO
Citations
5
Venue
IEEE International Conference on Robotics and Automation
Last Checked
6 months ago
Abstract
Multi-Agent Path Finding (MAPF) is the problem of effectively finding efficient collision-free paths for a group of agents in a shared workspace. The MAPF community has largely focused on developing high-performance heuristic search methods. Recently, several works have applied various machine learning (ML) techniques to solve MAPF, usually involving sophisticated architectures, reinforcement learning techniques, and set-ups, but none using large amounts of high-quality supervised data. Our initial objective in this work was to show how simple large scale imitation learning of high-quality heuristic search methods can lead to state-of-the-art ML MAPF performance. However, we find that, at least with our model architecture, simple large scale (700k examples with hundreds of agents per example) imitation learning does \textit{not} produce impressive results. Instead, we find that by using prior work that post-processes MAPF model predictions to resolve 1-step collisions (CS-PIBT), we can train a simple ML MAPF model in minutes that dramatically outperforms existing ML MAPF policies. This has serious implications for all future ML MAPF policies (with local communication) which currently struggle to scale. In particular, this finding implies that future learnt policies should (1) always use smart 1-step collision shields (e.g. CS-PIBT), (2) always include the collision shield with greedy actions as a baseline (e.g. PIBT) and (3) motivates future models to focus on longer horizon / more complex planning as 1-step collisions can be efficiently resolved.
Community Contributions
Found the code? Know the venue? Think something is wrong? Let us know!
π Similar Papers
In the same crypt β Multiagent Systems
R.I.P.
π»
Ghosted
R.I.P.
π»
Ghosted
Mean Field Multi-Agent Reinforcement Learning
π
π
The Cartographer
A Survey and Critique of Multiagent Deep Reinforcement Learning
π
π
The Cartographer
A Survey of Learning in Multiagent Environments: Dealing with Non-Stationarity
π
π
The Cartographer
Collaborative vehicle routing: a survey
R.I.P.
π»
Ghosted
Deep Reinforcement Learning for Swarm Systems
Died the same way β π» Ghosted
R.I.P.
π»
Ghosted
Federated Learning: Strategies for Improving Communication Efficiency
R.I.P.
π»
Ghosted
In-Datacenter Performance Analysis of a Tensor Processing Unit
R.I.P.
π»
Ghosted
Deep Convolutional Neural Networks for Computer-Aided Detection: CNN Architectures, Dataset Characteristics and Transfer Learning
R.I.P.
π»
Ghosted