Simulation-based Learning of Electrical Cabinet Assembly Using Robot Skills
Authors: Arik Laemmle, Balázs András Bálint, Philipp Tenbrock, Frank Naegele, David Traunecker, József Váncza, Marco F. Huber
Organizations: formerly Fraunhofer Institute for Manufacturing Engineering and Automation IPA, Stuttgart, Germany · formerly Institut für Strahlwerkzeuge IFSW, University of Stuttgart, Stuttgart, Germany · Research Laboratory on Engineering and Management Intelligence, Institute for Computer Science and Control SZTAKI, Budapest, Hungary · Institute of Industrial Manufacturing and Management IFF, University of Stuttgart, Fraunhofer Institute for Manufacturing Engineering and Automation IPA, Stuttgart, Germany
This paper presents a simulation-driven approach for automating the force-controlled assembly of electrical terminals on DIN-rails, a task traditionally hindered by high programming effort and product variability. The proposed method integrates deep reinforcement learning (DRL) with parameterizable robot skills in a physics-based simulation environment. To realistically model the snap-fit assembly process, we develop and evaluate two types of joining models: analytical models based on beam theory and rigid-body models implemented in the MuJoCo physics engine. These models enable accurate simulation of interaction forces, essential for training DRL agents. The robot skills are structured using the pitasc framework, allowing modular, reusable control strategies. Training is conducted in simulation using Soft Actor-Critic (SAC) and Twin Delayed Deep Deterministic Policy Gradient (TD3) algorithms. Domain randomization is applied to improve robustness. The trained policies are transferred to a physical UR10e robot system without additional tuning. Experimental results demonstrate high success rates (up to 100%) in both simulation and real-world settings, even under significant positional and rotational deviations. The system generalizes well to new terminal types and positions, significantly reducing manual programming effort. This work highlights the potential of combining simulation-based learning with modular robot skills for flexible, scalable automation in small-batch manufacturing. Future work will explore hybrid learning methods, automated environment parameterization, and further refinement of joining models for design integration.
Figures & tables
Figure 1: Structure of the electrical terminal assembly task, used in the presented work.
Figure 2: Proposed framework for learning robot skill parameters from simulation.
Figure 3: Unilaterally clamped bending beam and snap-hook (own illustration following [ 2 ] ).
Figure 4: Geometric relations (left) and acting forces (right) on the snap-hook head.
Figure 5: Rigid body joining model with one translational degree of freedom.
Figure 6: Rigid body joining model with one rotational degree of freedom.
Figure 7: Rigid body joining model with two rotational degrees of freedom.
Figure 8: Experimental setup for measuring the acting forces on the snap-hook.
Figure 9: Results from the experimental model evaluation and force progression for a snap-hook with Geometry II.
Figure 10: Lateral force (top) and joining force (bottom) values as recorded in the real-world scenario ( reference ) and provided by the rigid body joining models (1-DoF translational model " slide ", 1-DoF rotational model " one-hinge ", 2-DoF rotational model " two-hinge ") and the overlap-based extensions DRM ( measured ) and TFM ( calculated ). The vertical dashed line marks the point of the snap-in, i.e., the point until the models are considered valid.
Figure 11: Simulated model of the robotic test-bed including the integrated joining models.
Figure 12: Average return, success rate and number of used skills from the simulation-based training of a policy trained with SAC.
Figure 13: Success rates for a policy trained and evaluated on the WAGO 2002-1201 terminal during in-silico execution.
Policy
Trained for
Evaluated on
Experiment I
WAGO 2002-1201
WAGO 2002-1201
Experiment II
WAGO 2102-1301
WAGO 2102-1301
Experiment III
WAGO 2102-1301
WAGO 2016-1301
Table 1: Overview of trained and evaluated policies on the physical test-bed.
Figure 14: Success rates for a policy trained and evaluated on the WAGO 2002-1201 terminal during real-world execution.
Contact-rich robotic skills remain challenging for industrial robots due to tight geometric tolerances, frictional variability, and uncertain contact dynamics, particularly when using position-controlled manipulators. This paper presents a reusable and encapsulated skill-based strategy for peg-in-hole assembly, in which adaptation is achieved through Residual Reinforcement Learning (RRL). The assembly process is represented using composite skills with explicit pre-, post-, and invariant conditions, enabling modularity, reusability, and well-defined execution semantics across task variations. Safety and sample efficiency are promoted through RRL by restricting adaptation to residual refinements within each skill during contact-rich interactions, while the overall skill structure and execution flow remain invariant. The proposed approach is evaluated in MuJoCo simulation on a UR5e robot equipped with a Robotiq gripper and trained using SAC and JAX. Results demonstrate that the proposed formulation enables robust execution of assembly skills, highlighting its suitability for industrial automation.
Khalil Abuibaid, Aleksandr Sidorenko, Achim Wagner +1
Chair of Machine Tools and Control System, RPTU University Kaiserslautern-Landau, 67663 Kaiserslautern, Germany · Innovative Factory Systems, German Institute of Artificial Intelligence, 67663 Kaiserslautern, Germany · Chair of Machine Tools and Control Systems, RPTU University Kaiserslautern-Landau, 67663 Kaiserslautern, Germany
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.
A hybrid hierarchical learning framework is proposed to achieve high-precision assembly of 170mm cylindrical components with tolerance of 0.1mm. The lower-level network integrates expert experience through Behavior Cloning (BC), giving the robot human-like intuition, and incorporates the Twin Delayed Deep Deterministic Policy Gradient (TD3) algorithm to enhance training stability and robustness. The upper-level network dynamically adjusts the lower-level decisions based on heuristic rules, ensuring flexibility in operations. A simulated model is constructed to learn before transferring to real world. An efficient and safe training is allowed. Comparisons show that the reward curve converges within 500 episodes, indicating high learning efficiency. It also demonstrates better adaptability to initial conditions and pose errors, achieving satisfactory success rates even under extreme conditions. Moreover, the method exhibits good stability under Gaussian noise interference. In the real world, the assembly trajectory of the cylindrical segment shows smoother motion and less fluctuation.
Binbin Lian, Xinyu Liu, Tao Sun
Key Laboratory of Mechanism Theory and Equipment Design of Ministry of Education, Tianjin University, Tianjin, 300350, China