Co^2Gen: Collaboratively Generating Task Plans for Decentralized Multi-Robot and Human Collaboration

Yi Loo; Siddhant Shrivastava; Minh Long Pham; Malika Meghjani

cs.ROOct 22, 2026

Co^2Gen: Collaboratively Generating Task Plans for Decentralized Multi-Robot and Human Collaboration

Authors: Yi Loo, Siddhant Shrivastava, Minh Long Pham, Malika Meghjani

Organizations: Singapore University of Technology and Design (SUTD), Singapore.

Abstract

Visual Language Action (VLA) models are designed to be generalist robot models that can be implemented across embodiments, tasks and environments. However, their ability to collaborate with other robots and humans remains underexplored. In this work, we propose a framework to automate the generation of multi-robot collaborative data for decentralized VLA training without requiring human teleoperation. Our framework first utilizes a decentralized collaborative planning approach to generate multi-robot collaborative robot plans in the form of an executable function. We then utilize the collaborative plan to produce collaborative demonstration data to finetune VLA models in a decentralized manner. We show that our approach produces demonstration data that when used to finetune VLA models on decentralized collaboration tasks matches the task performance of VLA models finetuned on human-teleoperated collaborative data.

Figures & tables

Fig. 1: In decentralized task execution for a multi-robot setup, VLAs that are not trained on collaborative data fail to coordinate due to the lack of learned coordination behaviour.

Fig. 2: Our bi-manual arm setup for all of collaborative tasks.

Fig. 3: The system diagram for our proposed collaborative plan generation approach.

Fig. 5: Our framework for Collaborative Data Collection with a Human Verifier

Fig. 6: The two different collaborative tasks we evaluate our method on. Each of the two tasks require both arms to coordinate their sub-task execution in order to achieve overall task success.

Collaborative Stacking Task (Max Score $=14$ )
Sub-Task	Score
Arm 1 successfully picks up cube from its vicinity	$2$
Arm 2 successfully picks up cube from its vicinity	$2$
Arm 1 successfully places up the bottom cube at the central region between the arms	$2$
Arm 1 moves clears of the central region after placing the bottom cube	$3$
Arm 2 successfully stacks the top on the bottom cube	$5$

TABLE I: Sub-Tasks for Each Collaborative Task. Arm

i

denotes the Arm that takes on Role

i

of the task. A Sub-task is considered a success if the robot arm executes it as described without any collision with the other arm.

Fig. 7: Results for the Collaborative Storage and Stacking Tasks across 20 task executions. (a) Overall task score (TS) as scored in Table I . (b), (c) Sub-Task success rates (ST

x

SR) and overall task success rate (TSR) for the Storage and Stacking Tasks respectively. Error bars denote the standard error.

Fig. 8: Results for Human-Robot Collaboration on the Collaborative Storage and Stacking Tasks for each role taken by the robot arm. (a) Task success rate (TSR). (b) Role specific task score (TS).

Synthetic Data Generation for Robotics

Synthetic data generation for robotics creates simulated scenes, sensor readings, and action trajectories to train and evaluate robots.

See all 301 papers

CHORUS: Decentralized Multi-Embodiment Collaboration with One VLA Policy
Jun 10, 2026Ria Doshi, Tian Gao, Annie Chen +2Mobile Robotics Cross-Embodiment Robot Learning
Stanford University
MA-VLA: Multi-Arm Vision-Language-Action Model for Collaboration and Compositional Generalization
Aug 26, 2026Zaibin Zhang, Junlan Xiao, Zhongbo Zhang +11Vision-Language-Action Models Instruction Generalization in VLA Models
Dalian University of Technology · Shanghai Jiao Tong University · The Chinese University of Hong Kong, Shenzhen +4
D-VLC: Decentralized Vision-Language Collaboration for Heterogeneous Embodied Multi-Robot Systems in Unknown Environments
Jul 31, 2026Yuan Zhou, Ruitong Lin, Shen Wang +5Multi-Agent LLM Systems Multi-Robot Systems
Zhejiang University, Hangzhou, China · Differential Robotics, Hangzhou, China · Beihang University, Beijing, China

Jun 10, 2026J/K move · Enter open · S save

CHORUS: Decentralized Multi-Embodiment Collaboration with One VLA Policy

Ria Doshi, Tian Gao, Annie Chen, Chelsea Finn+1

Stanford University

Mobile Robotics Cross-Embodiment Robot Learning Multi-Robot Coordination Vision-Language-Action Models

Open paper AskPDF arXiv

Multi-robot collaboration allows robots to efficiently take on a wide range of tasks, from moving a couch through a doorway to assembling structures on a construction site. However, achieving such coordination in mobile multi-robot settings remains challenging: centralized methods conditioned on the combined observations of a team scale poorly with team size, and decentralized methods that train one policy per robot often require explicit alignment procedures or information sharing at inference time to overcome partial observability. Our key insight is that the visuomotor priors of pretrained vision-language-action (VLA) models should enable reactive, decentralized collaboration from each robot's local observations alone, without these inference-time assumptions. We propose CHORUS, a framework that adapts a single VLA backbone to control diverse, multi-robot teams. At inference time, each robot runs an independent copy of CHORUS, conditioned only on its own observations and a robot-identifying prompt. In real-world experiments including mobile tape measurement, library book handovers, and laundry basket lifting, CHORUS achieves a 64% point improvement over decentralized, from-scratch models, improves reactivity to teammate behavior by 40% points, and outperforms centralized baselines. Together, these results show that a shared VLA backbone is capable of achieving decentralized multi-robot collaboration, without per-robot policies or inter-robot communication at inference.

Abstract

Synthetic Data Generation for Robotics

Explore similar work

CHORUS: Decentralized Multi-Embodiment Collaboration with One VLA Policy

MA-VLA: Multi-Arm Vision-Language-Action Model for Collaboration and Compositional Generalization

D-VLC: Decentralized Vision-Language Collaboration for Heterogeneous Embodied Multi-Robot Systems in Unknown Environments