Towards simulation-based optimization of compliant fingers for high-speed connector assembly
Richard Matthias Hartisch, Alexander Rother, Jörg Krüger, Kevin Haninger · 2025
Read Towards simulation-based optimization of compliant fingers for high-speed connector assembly by Richard Matthias Hartisch, Alexander Rother, Jörg Krüger, Kevin Haninger (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2509-10012.txt.
Introduction
Richard Matthias Hartisch and 3 co-authors optimise soft robot fingers inside a simulator in a 2025 robotics paper aimed at high-speed connector assembly. A finger's give is a design choice that decides whether a plug seats cleanly or jams when parts arrive slightly off position. Trying shapes on real hardware is slow, and flat simplified models cannot score a whole insertion task, so the team designs against task success inside dynamic contact simulation instead.
Tuning the structured finger's geometry and stiffness in that virtual bench widened the window of starting positions the robot could tolerate, more than doubling it. The winning design was then built and checked on a real robot using standard benchmark board tasks. Simulation-led tuning of softness, the authors argue, is ready for contact-rich work.
Catalog record
| CREATOR | Richard Matthias Hartisch, Alexander Rother, Jörg Krüger, Kevin Haninger |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 40 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2509-10012
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Towards simulation-based optimization of compliant fingers for high-speed connector assembly Richard Matthias Hartisch, Alexander Rother, Jörg Krüger, Kevin Haninger Abstract Mechanical compliance is a key design parameter for dynamic contact-rich manipulation, affecting task success and safety robustness over contact geometry variation. Design of soft robotic structures, such as compliant fingers, requires choosing design parameters which affect geometry and stiffness, and therefore manipulation performance and robustness. Today, these parameters are chosen through either hardware iteration, which takes significant development time, or simplified models (e.g. planar), which can’t address complex manipulation task objectives. Improvements in dynamic simulation, especially with contact and friction modeling, present a potential design tool for mechanical compliance. We propose a simulation-based design tool for compliant mechanisms which allows design with respect to task-level objectives, such as success rate. This is applied to optimize design parameters of a structured compliant finger to reduce failure cases inside a tolerance window in insertion tasks. The improvement in robustness is then validated on a real robot using tasks from the benchmark NIST task board. The finger stiffness affects the tolerance window: optimized parameters can increase tolerable ranges by a factor of 2.29 2.29 , with workpiece variation up to 8.6 m m 8.6\ mm being compensated. However, the trends remain task-specific. In some tasks, the highest stiffness yields the widest tolerable range, whereas in others the opposite is observed, motivating need for design tools which can consider application-specific geometry and dynamics. Simulation for optimization of compliant fingers 1 1 footnotetext: Department of Industrial Automation Technology at TU Berlin, Germany. 2 2 footnotetext: Department of Automation at Fraunhofer IPK, Berlin, Germany. † † footnotetext: Corresponding author: [email protected] . This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No. 101058521 — CONVERGING. 1 Introduction Robust contact-rich manipulation requires compensation for small errors in the contact geometry [ 1 , 2 ] . Mechanical compliance offers compensation of variation in contact geometry through intrinsic dynamics; reducing the sensitivity to position variations and enabling self-correcting intrinsic dynamics, e.g. self-centering [ 3 , 4 ] . The importance of compliance can be seen from its robotic applications: integrated in the kinematic structure [ 5 ] , joints [ 6 ] , flange [ 3 , 7 ] , fingers [ 8 , 1 ] , and environment [ 9 , 10 , 11 ] . Mechanical compliance directly affects normal forces and surface geometry during contact. This directly affects pressure distribution on contact areas, Coulomb friction limits, and the robustness uniformity over variation in object contact condition. The mechanical design of compliance, however, is limited to lower-dimensional, simplified models (e.g. planar systems) to enable design analysis such as constraint-based design [ 12 ] , effective TCP stiffness matrix [ 13 , 10 ] , or compliance ellipsoids [ 14 ] . While these methods can develop design intuition, they cannot scale to more complex designs. Topological optimization methods can improve the range of motion or linearity of compliant mechanisms [ 15 ] , but has not been investigated for task-related performance. Mechanical parameters do not have the bandwidth limitations of active control [ 1 ] , therefore methods for mechanical compliance design are considered here. As 3D printed compliant structures enables a larger design space, methods are needed to optimize these designs with respect to task-level performance: e.g. robustness over part variation, in-hand slip, maximum collision force. Figure 1: Comparison of current (top) and proposed (bottom) design pipeline of passive, structured compliance. The objective is to reduce the need for real-life experiments, using instead a robot model and task (enter on the left) and dynamic simulation to evaluate a range of design and environmental parameters. The simulation results are evaluated according to robustness over part pose variation and failure cases. In this paper, these simulation results are compared against real execution. At the same time, dynamic simulation has substantially improved for robotics in terms of geometric complexity it can handle, numerical stability of simulation, differentiability, and realism. This offers a potential design tool for compliance, one which can scale in task and design complexity as well as provide task-level evaluation. Soft systems, e.g. compliant grippers, are numerically easier to simulate compared to stiff systems, which may allow for a reduced sim2real gap [ 16 ] . However, the question of simulation accuracy remains critical: is simulation accurate enough to motivate the design decisions in compliant device design? Simulation has been used as a tool to iteratively optimize fingertip contact geometry and grasping success [ 17 ] , but not yet applied for compliant fingers. To investigate applicability, MuJoCo is chosen as a physics-engine that focuses on contact-rich tasks with sufficient accuracy, modeling the 6-DOF stiffness between object contact and robot flange, and investigating if the simulation can predict the tolerance window in a resulting assembly task. Additionally, the ability to determine failure cases, i.e. slip, a failed search strategy and jamming of the plug is investigated. An overview of the approach can be seen in Figure 1 . 2 Task Description This section describes the manipulation tasks considered, defining the assembly goal and failure cases. 2.1 Assembly Applications (a) KET12 (b) KET8 (c) RJ45 (d) USB-A Connector Figure 2: The assembly tasks considered in this paper are shown. Four assembly tasks have been chosen from the NIST task board, which are both simulated and carried out in real-life experiments. These include two rectangular shaped pegs in varying sizes (KET12 in Fig. 2a and KET8 in Fig. 2b ), as well as two electrical connectors (RJ45 in Fig. 2c and USB in Fig. 2d ). The assembly sequences utilized are similar for all assembly tasks and are further described in the simulation preparation in Sec 4.2 . 2.2 Robot Assembly Process To achieve high-speed insertion, compliant fingers are used along with an assembly strategy, based on the search strategy employed in [ 1 ] , which can can be seen in Fig. 3 . First, the robot is positioned in its home position. Next, the robot is moved from the home position to above the connector within a safe distance. Next, it is lowered to the gripping height, after which the grippers are closed, grasping the part. The robot then moves upwards again to the previous position, while holding on to the connector. The aim of this process is to ensure that the connector is aligned with the socket before the assembly is performed via search strategy. The process is identical for all assembly tasks, except the starting position and the distance that the robot has to move for alignment are different. (a) (b) (c) (d) (e) Figure 3: Here, the simulated assembly movement is demonstrated exemplary for the USB assembly task. (a) and (b) visualize the approaching movement and first established contact on the socket. In (c) and (d) a back-and-forth motion of the connector in the x- direction with a slight movement in the z- direction is executed to allow the plug to slip into the socket when aligned. Following, in (e) contact is established between the sides of the plug and socket, realized by a motion in the negative y- direction. 2.3 Failure Cases