MALLVI: A Multi-Agent Framework for Integrated Generalized Robotics Manipulation

TL;DR

MALLVI introduces a multi-agent framework combining vision and language models for robust, closed-loop robotic manipulation that improves success rates through iterative feedback and specialized agent coordination.

Contribution

It presents a novel multi-agent system that integrates perception, reasoning, and feedback for robotic manipulation, outperforming prior open-loop approaches.

Findings

Closed-loop multi-agent coordination enhances manipulation success.

The framework generalizes well in zero-shot settings.

Code is available at https://github.com/iman1234ahmadi/MALLVI.

Abstract

Task planning for robotic manipulation with large language models (LLMs) is an emerging area. Prior approaches rely on specialized models, fine tuning, or prompt tuning, and often operate in an open loop manner without robust environmental feedback, making them fragile in dynamic settings. MALLVI presents a Multi Agent Large Language and Vision framework that enables closed-loop feedback driven robotic manipulation. Given a natural language instruction and an image of the environment, MALLVI generates executable atomic actions for a robot manipulator. After action execution, a Vision Language Model (VLM) evaluates environmental feedback and decides whether to repeat the process or proceed to the next step. Rather than using a single model, MALLVI coordinates specialized agents, Decomposer, Localizer, Thinker, and Reflector, to manage perception, localization, reasoning, and high level planning. An optional Descriptor agent provides visual memory of the initial state. The Reflector supports targeted error detection and recovery by reactivating only relevant agents, avoiding full replanning. Experiments in simulation and real-world settings show that iterative closed loop multi agent coordination improves generalization and increases success rates in zero shot manipulation tasks. Code available at https://github.com/iman1234ahmadi/MALLVI .