Data to 14 Oct 2026

Harnessing human expertise for high-precision robotic assembly in industrialized construction: A sample-efficient installer-in-the-loop interactive reinforcement learning framework

arXiv 2609.13234 ↗First posted 2 Sep 2026cs.RO stat.MLcited by 0

On the map: Offline RL and imitation learning from data → Real-world robot offline RL

also in 1 more

Deep reinforcement learning methods → RL for robot locomotion and manipulation

Abstract (arXiv)

Industrialized construction imposes stringent precision requirements on robotic assembly of modular components such as prefabricated window units. In tolerance-critical operations, the central bottleneck is not only mechanical clearance but also converting tacit installer expertise into data-efficient autonomy under sparse acceptance feedback, contact variability, and millimeter-scale constraints. We present an installer-in-the-loop interactive reinforcement learning framework that acquires expertise through offline teleoperated demonstrations, sparse event-driven binary takeovers at contact-failure boundaries, and acceptance-aligned terminal rewards, logged under a unified schema for traceable offline-to-online adaptation. A temporally abstract action-sequence policy built on Q-chunking with Flow Q-Learning captures multimodal recovery maneuvers under sparse terminal rewards, while a non-updating warm-start phase stabilizes the offline-to-online transition. The framework is evaluated in MuJoCo across the workflow from suction acquisition through clearance-limited seating, under structured staging and end-to-end randomized placement. Within a defined stress-test regime with 2 mm per-side clearance, bounded pose perturbations, and friction randomization, the pipeline attains 100% autonomous seating with 12--15 min of cumulative installer supervision over 3.0 h of online training, and reaches the 95% success milestone in approximately 0.5 h and 1.5 h in the two experiments. We also report wall-clock adaptation time, cumulative takeover minutes, intervention-rate decay, and stage-wise failure attribution to inform supervision budgeting. Ablations isolate the complementary contributions of temporal abstraction, installer intervention, and warm-start value calibration.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Real-world robot offline RL, one of 13 directions in Offline RL and imitation learning from data. Builds on 34 of its 71 references that are in the map. First posted on arXiv 2 Sep 2026.

Builds on

34 of 71 references.

References · 34

2015Deep Residual Learning for Image Recognitioncited by 52,096
2018Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actorcited by 4,885
2022Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flowcited by 3,700
2023Diffusion Policy: Visuomotor Policy Learning via Action Diffusioncited by 2,838
2023Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardwarecited by 1,568
2017Hindsight Experience Replaycited by 1,136
2015End-to-End Training of Deep Visuomotor Policiescited by 875
2018QT-Opt: Scalable Deep Reinforcement Learning for Vision-Based Robotic Manipulationcited by 743
2017Learning Complex Dexterous Manipulation with Deep Reinforcement Learning and Demonstrationscited by 715
2017Sim-to-Real Transfer of Robotic Control with Dynamics Randomizationcited by 672
2023Efficient Online Reinforcement Learning with Offline Datacited by 363
2016Deep Reinforcement Learning for Robotic Manipulation with Asynchronous Off-Policy Updatescited by 353

22 more

2017Leveraging Demonstrations for Deep Reinforcement Learning on Robotics Problems with Sparse Rewardscited by 319
2018Residual Reinforcement Learning for Robot Controlcited by 274
2018HG-DAgger: Interactive Imitation Learning with Human Expertscited by 266
2018Safe Exploration in Continuous Action Spacescited by 247
2018Closing the Sim-to-Real Loop: Adapting Simulation Randomization with Real World Experiencecited by 217
2024SERL: A Software Suite for Sample-Efficient Robotic Reinforcement Learningcited by 136
2017Stochastic Neural Networks for Hierarchical Reinforcement Learningcited by 124
2020Human-in-the-Loop Imitation Learning using Remote Teleoperationcited by 112
2017Trial without Error: Towards Safe Reinforcement Learning via Human Interventioncited by 103
2019Deep Reinforcement Learning for Industrial Insertion Tasks with Visual Inputs and Natural Rewardscited by 64
2015Reinforcement Learning with Parameterized Actionscited by 63
2024Efficient Online Reinforcement Learning Fine-Tuning Need Not Retain Offline Datacited by 63
2022Factory: Fast Contact for Robotic Assemblycited by 54
2021A Survey on Interactive Reinforcement Learning: Design Principles and Open Challengescited by 41
2017Deep Reinforcement Learning for High Precision Assembly Taskscited by 39
2015Learning Contact-Rich Manipulation Skills with Guided Policy Searchcited by 38
2021Robust Multi-Modal Policies for Industrial Assembly via Reinforcement Learning and Demonstrations: A Large-Scale Studycited by 30
2025Actor-Critic for Continuous Action Chunks: A Reinforcement Learning Framework for Long-Horizon Robotic Manipulation with Sparse Rewardcited by 13
2021Interpreting Contact Interactions to Overcome Failure in Robot Assembly Taskscited by 9
2024Symmetry-aware Reinforcement Learning for Robotic Assembly under Partial Observability with a Soft Wristcited by 9
2024ExACT: An End-to-End Autonomous Excavator System Using Action Chunking With Transformerscited by 4
2023Sensory Manipulation as a Countermeasure to Robot Teleoperation Delays: System and Evidencecited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way