Toward Deployable Multi-Robot Collaboration via a Symbolically-Guided Decision Transformer

TL;DR

This paper introduces SGDT, a hierarchical neuro-symbolic framework that combines symbolic planning with decision transformers to enable effective, interpretable, and deployable multi-robot collaboration in complex tasks.

Contribution

It presents the first application of decision transformers to multi-robot manipulation, integrating symbolic planning with deep learning for structured decision making.

Findings

SGDT achieves effective multi-robot collaboration in various scenarios.

The hierarchical approach improves interpretability and generalization.

SGDT outperforms baseline methods in zero-shot and few-shot tasks.

Abstract

Reinforcement learning (RL) has demonstrated great potential in robotic operations. However, its data-intensive nature and reliance on the Markov Decision Process (MDP) assumption limit its practical deployment in real-world scenarios involving complex dynamics and long-term temporal dependencies, such as multi-robot manipulation. Decision Transformers (DTs) have emerged as a promising offline alternative by leveraging causal transformers for sequence modeling in RL tasks. However, their applications to multi-robot manipulations still remain underexplored. To address this gap, we propose a novel framework, Symbolically-Guided Decision Transformer (SGDT), which integrates a neuro-symbolic mechanism with a causal transformer to enable deployable multi-robot collaboration. In the proposed SGDT framework, a neuro-symbolic planner generates a high-level task-oriented plan composed of symbolic subgoals. Guided by these subgoals, a goal-conditioned decision transformer (GCDT) performs low-level sequential decision-making for multi-robot manipulation. This hierarchical architecture enables structured, interpretable, and generalizable decision making in complex multi-robot collaboration tasks. We evaluate the performance of SGDT across a range of task scenarios, including zero-shot and few-shot scenarios. To our knowledge, this is the first work to explore DT-based technology for multi-robot manipulation.