RoCo Challenge at AAAI 2026: Benchmarking Robotic Collaborative Manipulation for Assembly Towards Industrial Automation
Authors: Haichao Liu, Yuheng Zhou, Zhenyu Wu, Ziheng Ji, Ziyu Shan, Qianzhun Wang, Ruixuan Liu, Zhiyuan Yang, Yejun Gu, Shalman Khan, Shijun Yan, Jun Liu, Haiyue Zhu, Changliu Liu, Jianfei Yang, Jingbing Zhang, Ziwei Wang
Organizations: Nanyang Technological University, Singapore · Beijing University of Posts and Telecommunications, Beijing, China · Carnegie Mellon University, Pittsburgh, USA · Agency for Science, Technology and Research (A*STAR), Singapore
Embodied Artificial Intelligence (EAI) is rapidly developing, gradually subverting previous autonomous systems' paradigms from isolated perception to integrated, continuous action. This transition is highly significant for industrial robotic manipulation, promising to free human workers from repetitive, dangerous daily labor. To benchmark and advance this capability, we introduce the Robotic Collaborative Assembly Assistance (RoCo) Challenge with a dataset towards simulation and real-world assembly manipulation. Set against the backdrop of human-centered manufacturing, this challenge focuses on a high-precision planetary gearbox assembly task, a demanding yet highly representative operation in modern industry. Built upon a self-developed data collection, training, and evaluation system in Isaac Sim, and utilizing a dual-arm robot for real-world deployment, the challenge operates in two phases. The Simulation Round defines fine-grained task phases for step-wise scoring to handle the long-horizon nature of the assembly. The Real-World Round mirrors this evaluation with physical gearbox components and high-quality teleoperated datasets. The core tasks require assembling an epicyclic gearbox from scratch, including mounting three planet gears, a sun gear, and a ring gear. Attracting over 60 teams and 170+ participants from more than 10 countries, the challenge yielded highly effective solutions, most notably ARC-VLA and RoboCola. Results demonstrate that a dual-model framework for long-horizon multi-task learning is highly effective, and the strategic utilization of recovery-from-failure curriculum data is a critical insight for successful deployment. This report outlines the competition setup, evaluation approach, key findings, and future directions for industrial EAI. Our dataset, CAD files, code, and evaluation results can be found at: https://rocochallenge.github.io/RoCo2026/.
Figures & tables
Figure 1: The simplified planetary gear reducer used in the RoCo Challenge, consisting of a sun gear, three planet gears, one finned key shaft, a ring gear component, and a planet carrier with three pins. The assembly task requires precise placement of these components within a planet carrier.
Figure 2: Simulation environment demonstrations for the three RoCo Challenge tasks. The stages mirror the real-world setup: (a) initial assembly, (b) completion from partial states, and (c) dynamic error recovery.
Figure 3: Visualization of the data frame for the simulation track. The dataset includes synchronized multi-modal observations (RGB, Depth, and proprioception) and action streams, enabling comprehensive training for both visuomotor imitation learning and VLA models.
Figure 4: Hardware settings for real-world data collection. The platform consists of a Galaxea A1X robot arm, multiple RealSense D405 cameras, a binocular camera, and an R1 Lite teleoperation device used for demonstration collection.
Figure 5: Real-world demonstration of the three gear assembly tasks in the RoCo Challenge. (a) Assembly from scratch; (b) Completion from a partial state; (c) Error recovery by replacing the incorrect gear.
Figure 6: Visualization of the data frame for the real-world track collected using the teleoperation device. The dataset includes synchronized multi-modal observations (RGB, Depth, and proprioception) and action streams, enabling comprehensive training in real-world conditions.
Figure 7: Overall pipeline of ARC-VLA. The model integrates task instruction, multi-camera observations, and robot state features through a multi-modal backbone and a failure-aware policy learning framework to generate closed-loop manipulation actions.
Figure 8: Overall pipeline of the Team RoboCola. The framework adopts a dual-model collaboration strategy to couple high-level task reasoning with low-level motion generation, enabling effective and precise robotic manipulation.
Robot learning has undergone significant advancements in the domain of compliant control and robotic assembly. The development of novel methodologies has transcended the capacity to address discrete tasks, such as grasping or inserting, and has evolved to encompass the comprehensive construction of entire assemblies. Conventional benchmarks, which are predominantly focused on single tasks, are no longer adequate to address the complexity and interconnected nature of modern manufacturing processes. Consequently, the necessity for benchmarks to encompass a broader spectrum of assembly and subassembly steps is becoming apparent. In light of these developments, we propose LABIT, an industrial robotic assembly benchmark for long-horizon manipulation. This benchmark encompasses DIN-norm assembly tasks, including screwing, placing, inserting, and snapping, under both high and low tolerance conditions. The successive nature of these assembly steps implies that prior errors can influence subsequent manipulation. Consequently, this benchmark is particularly relevant for assessing the robustness of intelligent robotic systems. The proposed benchmarking protocols and novel performance metric, based on a taxonomy of assembly complexity in conjunction with the comprehensive list of parts and their corresponding CAD models, serve as essential tools for research and development in the field. To provide a baseline, we evaluate the benchmark using traditional force-position control and report reliability and score metrics per primitive, subtask, assembly category, and subassembly. Accompanying data is available at https://irp-hka.github.io/robotic-assembly-benchmark/.
Constantin Schempp, Christian Friedrich, Luca Kofler +1
Institute for Robotics and Intelligent Production Systems, University of Applied Sciences Karlsruhe (HKA), Moltkestraße 30, Karlsruhe, 76133, Germany
Dexterous manipulation remains a critical bottleneck in industrial automation; tasks such as cable routing, connector insertion, and precision assembly still rely heavily on manual labor despite decades of robotics research. This work presents a progression from classical, modular robotics pipelines toward an end-to-end multimodal imitation-learning framework for industrial dexterous manipulation. As a part of this work, we introduce three key contributions: a set of Industrial Dexterity Benchmark (IDB) boards aimed to mimic datacenter cable management, automotive cable harnesses, and gearbox assembly tasks; a scalable imitation learning framework (DAG-ROS); and a multimodal diffusion-based policy framework (AG-iDP3) that creates models fusing RGB images, point clouds, joint positions, and wrist-frame wrench data. Focusing on the datacenter cable manipulation board, we evaluate the performance of a task involving cleaning a single cable over variations of an end-to-end AI policy using 48 trials per configuration. The best performing configuration, a multimodal expansion Diffusion Policy (DP), includes a multi-view RGB image source passed through an R3M encoder and reaches a 78% grasp and insert combined task success rate. This performance marks a significant improvement over the 36% observed from the single-camera RGB DP baseline. Each of the tested configurations requires only approximately 100 teleoperated demonstrations per task phase. These results indicate that the correct learned policy can outperform classical vision and control robotic methods in robustness, generalization, and deployment efficiency, justifying a shift toward scalable robotic automation for high up-time industrial environments.
Robotic assembly presents a long-standing challenge due to its requirement for precise, contact-rich manipulation. While simulation-based learning has enabled the development of robust assembly policies, their performance often degrades when deployed in real-world settings due to the sim-to-real gap. Conversely, real-world reinforcement learning (RL) methods avoid the sim-to-real gap, but rely heavily on human supervision and lack generalization ability to environmental changes. In this work, we propose a hybrid approach that combines a simulation-trained base policy with a real-world residual policy to efficiently adapt to real-world variations. The base policy, trained in simulation using low-level state observations and dense rewards, provides strong priors for initial behavior. The residual policy, learned in the real world using visual observations and sparse rewards, compensates for discrepancies in dynamics and sensor noise. Extensive real-world experiments demonstrate that our method, SPARR, achieves near-perfect success rates across diverse two-part assembly tasks. Compared to the state-of-the-art zero-shot sim-to-real methods, SPARR improves success rates by 38.4% while reducing cycle time by 29.7%. Moreover, SPARR requires no human expertise, in contrast to the state-of-the-art real-world RL approaches that depend heavily on human supervision.