To address the inherent nondeterminism and lack of verifiability of learning- or sampling-based path planning methods in safety-critical dual-robot cooperative manufacturing, this paper aims to develop a deterministic and decoupled kinematic framework that ensures submillimeter accuracy, repeatability and native compatibility with industrial robot controllers for tasks such as precision drilling and riveting of large aerospace components.
The proposed framework is founded on first principles of geometry. For cooperative following, an extended tool frame method is introduced to transform complex coupled motions into independent single-robot trajectories executable by standard robot commands. For symmetric tasks, an analytical mirroring algorithm based on an arbitrary spatial plane is developed to offline generate accurate follower trajectories. The approach relies on high-precision laser tracker calibration and is validated on a dual-KUKA KR500-3 industrial robot platform.
Experimental validation demonstrates that the framework achieves a total cooperative positioning error of less than 0.53 mm under realistic industrial conditions. Both cooperative following and cooperative mirroring tasks satisfy the stringent tolerance requirements of aerospace manufacturing, confirming submillimeter accuracy, high repeatability and practical feasibility on heavy-duty industrial robots.
This work presents a novel deterministic kinematic framework that decouples dual-robot cooperative constraints into single-robot tracking problems without modifying low-level controller code or relying on real-time external computation. The extended tool frame and analytical mirroring methods significantly reduce programming effort while guaranteeing path determinism and verifiability. The submillimeter accuracy achieved meets aerospace-grade standards, offering a practical, efficient and certifiable solution for high-precision dual-robot cooperative manufacturing.
