Robotics · 2026-10-14
Akule: Fast and Scalable Multi-Robot Motion Planning via Sparse Interaction Diffusion
Manav Vora, Gokul Puthumanaillam, Aditya Penumarti, Jane Shin, Melkior Ornik
AI-assisted summary based on the paper. Consult the linked source for methods and limitations. Benchmark numbers appear after editor verification.
TL;DR
Akule augments a frozen single-robot diffusion prior for multi-robot motion planning with (1) a reusable pairwise correction model Rϕ, (2) a contextual selector Uψ that chooses which directed pairwise corrections to evaluate sparsely, and (3) a composer Gω that combines selected corrections using context-dependent signed weights. Joint proposals are passed to downstream hard repair. The authors report competitive success, up to 25× speedups, zero-shot population transfer, and scalability to 256 robots.
Why it matters
Scaling diffusion-based MRMP to large teams requires coordinating robot trajectories without expensive dense interaction computation. By separating (a) which interactions to compute from (b) how to compose their effects, Akule aims to reduce costly pairwise evaluations while still producing proposals that downstream search can validate and repair. The paper also emphasizes permutation-equivariant design and learning without requiring a ground-truth interaction graph.
Method
- Frozen unary diffusion prior produces independent intent; Akule adds sparse interaction reasoning during each reverse diffusion step.
- A contextual selector Uψ predicts a directed interaction mask E_t to decide which ordered robot pairs (i,j) to evaluate with the reusable pairwise correction model Rϕ.
- A composer Gω assigns context-dependent signed coefficients α_ij,t to combine selected corrections into a coordinated denoising prediction before downstream conflict-based search repair.
Limitation
Candidate scoring: the selector evaluates directed candidates before hard admission, so its pointwise descriptors and contextual processing remain quadratic in robot count, even when pairwise corrections are evaluated sparsely. Savings in Rϕ evaluations do not imply a linear-time complete sampler or planner.
Abstract (from arXiv)
Diffusion models provide expressive trajectory priors for multi-robot motion planning (MRMP), but scaling them to large teams requires coordinating individual motions without expensive joint computation. We introduce Akule, a sparse interaction diffusion framework that augments a frozen single-robot prior with three learned modules: a reusable pairwise correction model $R_{\phi}$, a contextual selector $U_{\psi}$ that determines which corrections to evaluate, and a composer $G_{\omega}$ that combines their effects through context-dependent signed weights. This decomposition separates which interactions merit computation from how their effects should be combined into coordinated denoising updates. We train $R_{\phi}$ using matched pairwise trajectory optimizations, then learn $G_{\omega}$ and $U_{\psi}$ from physically optimized joint trajectories without requiring a prescribed ground-truth interaction graph. The resulting joint proposals are passed to a downstream search-based planner for validation and repair. Across diverse MRMP benchmarks, Akule achieves competitive planning success, complete-planner speedups of up to $25\times$, zero-shot population transfer, and successful planning with up to $256$ robots. Ablations further demonstrate that learned selection and composition reduce pre-repair conflicts and downstream repair effort relative to heuristic selection and uniform summation.
Related papers
- Marginal-Preserving Coupled Diffusion Composition for Long-Horizon Planning
- Training-Free Diffusion Planning with Analytical Local Scores
- Identifiable World Models from Pretrained Diffusion Representations
- DPDP: Diffusion Policy-Steered Diffusion Planner for Offline Reinforcement Learning
- HiLo-WAM: Decoupling Long-Horizon Task Evolution and Local Dynamics in World Action Models
Publication record
Bibliographic evidence describes where and when a record appeared. It does not validate scientific claims.
First public release: 2026-10-09 · Indexed by Pub-AI: 2026-10-14 UTC
Akule: Fast and Scalable Multi-Robot Motion Planning via Sparse Interaction Diffusion
Preprint posted: 2026-10-09 · Original record ↗
- arxiv
- 2610.14248
Explicit source versions (1)
- v1 · 2026-10-09
- Preprint available
Source date: 2026-10-09 · First detected: 2026-10-25 UTC · Historical baseline
Source evidence (1)
- Official evidence ↗ · atom:published (official metadata)
Abstract · arxiv
Diffusion models provide expressive trajectory priors for multi-robot motion planning (MRMP), but scaling them to large teams requires coordinating individual motions without expensive joint computation. We introduce Akule, a sparse interaction diffusion framework that augments a frozen single-robot prior with three learned modules: a reusable pairwise correction model $R_φ$, a contextual selector $U_ψ$ that determines which corrections to evaluate, and a composer $G_ω$ that combines their effects through context-dependent signed weights. This decomposition separates which interactions merit computation from how their effects should be combined into coordinated denoising updates. We train $R_φ$ using matched pairwise trajectory optimizations, then learn $G_ω$ and $U_ψ$ from physically optimized joint trajectories without requiring a prescribed ground-truth interaction graph. The resulting joint proposals are passed to a downstream search-based planner for validation and repair. Across diverse MRMP benchmarks, Akule achieves competitive planning success, complete-planner speedups of up to $25\times$, zero-shot population transfer, and successful planning with up to $256$ robots. Ablations further demonstrate that learned selection and composition reduce pre-repair conflicts and downstream repair effort relative to heuristic selection and uniform summation.
