A Unified Control Architecture for Macro-Micro Manipulation using a Active Remote Center of Compliance for Manufacturing Applications

TL;DR

The paper tackles the bandwidth bottleneck in macro-micro manipulation by integrating the macro manipulator into active interaction control via an Active Remote Center of Compliance architecture. It develops surrogate models and uses \mathcal{H}_\infty$ synthesis to design fixed-structure controllers that jointly govern macro and micro dynamics, achieving significant bandwidth gains over leader-follower and robot-based force control. Experimental results across force bandwidth identification, collision tasks, force-trajectory following, and assembly benchmarks demonstrate faster contact, improved force tracking, and reduced cycle times, with quantified improvements such as a $2.1$× and $12.5$× bandwidth increase and substantial time reductions in peg-in-hole, gear, and circuit-board assembly. The approach promises practical impact in manufacturing by enabling higher-speed, more sensitive contact tasks and easier hardware adaptation through surrogate models, with future work incorporating visual sensing to predict contact timing and environment behavior.

Abstract

Macro-micro manipulators combine a macro manipulator with a large workspace, such as an industrial robot, with a lightweight, high-bandwidth micro manipulator. This enables highly dynamic interaction control while preserving the wide workspace of the robot. Traditionally, position control is assigned to the macro manipulator, while the micro manipulator handles the interaction with the environment, limiting the achievable interaction control bandwidth. To solve this, we propose a novel control architecture that incorporates the macro manipulator into the active interaction control. This leads to a increase in control bandwidth by a factor of 2.1 compared to the state of the art architecture, based on the leader-follower approach and factor 12.5 compared to traditional robot-based force control. Further we propose surrogate models for a more efficient controller design and easy adaptation to hardware changes. We validate our approach by comparing it against the other control schemes in different experiments, like collision with an object, following a force trajectory and industrial assembly tasks.

Figures (14)

• Figure 1: (a) Enhanced module of the Active Remote Center of Compliance (ARCC) from friedrichHighlyDynamicPhysical2025. Each module consists of a BLDC-servodrive, a spindle, a flexure hinge and a position sensor to measure the spring deflection. (b) Configuration with three linear modules mounted on a industrial robot.
• Figure 2: Mechanical model of the macro-micro manipulator with the corresponding transfer functions. Index $M$ denotes the macro manipulator, index $\mu$ the micro manipulator with the additional indices $a$ for the active part and $p$ for the passive part, respectively.
• Figure 3: (a) Sampling positions for the system identification inside the expected workspace ($\mathrm{X{\times} Y{\times} Z}$) of $\unit[200{\times} 600{\times} 300]{mm}$. (b) Resulting boxplot of the identified cutoff frequencies $\omega_{cM}$ for the three linear DoFs of the macro manipulator.
• Figure 4: Bode plots of the identified transfer functions for the macro manipulator $G_M$, the micro manipulators active side $G_{\mu a}$ and passive side $G_{\mu p}$. The additional indices $low$ and $high$ denote the low stiffness and high stiffness flexure hinge, respectively. (a) X-axis with $k_{\mu, low}=15\;\unitfrac{N}{mm}$ and $k_{\mu, high}=30\;\unitfrac{N}{mm}$, (b) Y-axis with $k_{\mu, low}=15\;\unitfrac{N}{mm}$ and $k_{\mu, high}=30\;\unitfrac{N}{mm}$, (c) Z-axis with $k_{\mu, low}=20\;\unitfrac{N}{mm}$ and $k_{\mu, high}=40\;\unitfrac{N}{mm}$.
• Figure 5: (a) Sampling positions inside the expected workspace for the measurement of the environment stiffness $k_{env}$ used in the surrogate model described by (\ref{['eq:tfEnvironment']}). (b), (c), (d) Measured stiffnesses for the X-axis, Y-axis and Z-axis, respectively.