In context
Robotic manipulation of deformable objects—such as fabrics, cables, and soft tissues—has been a long-standing challenge in industrial automation. Traditional rigid-body robotics methods fail to model the high-dimensional, continuous deformation of these objects, limiting their use in tasks like cable harnessing in aerospace or soft tissue handling in surgery. By early 2024, the gap between accurate but slow finite element methods and fast but less accurate computer graphics simulations remained a key bottleneck.
What was reported
Researchers from the Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, introduced an implicit optimization method based on the alternating direction multiplier method (ADMM) to compute object deformation under robotic manipulation. The method takes a general geometric model as input, generates a finite element mesh, and constructs material constitutive potential energy and positional constraint energy functions for manipulation and collision interactions.
Using a projective dynamics strategy, the deformation problem is formulated as a two-term optimization solved via ADMM, enabling parallel computation on multi-threaded CPUs or GPUs. Numerical experiments demonstrated that the method achieves a physical deformation update rate exceeding 24 frames per second while keeping relative deformation error below 5%.
The approach was validated in practical scenarios, including quantitative evaluation from deformation simulation to online manipulation execution, as well as offline planning and simulation in globally constrained environments.
Why it mattered
This work bridges the gap between computational efficiency and physical accuracy, potentially enabling real-time model-based planning and control for deformable object manipulation in industrial settings. By providing a fast and accurate deformation solver, it opens the door for autonomous robots to handle flexible materials in manufacturing, logistics, and service applications.
“The proposed method can achieve an updating rate greater than 24 frames per second for physical deformation calculation while guaranteeing a relative deformation error less than 5%.”
Source: 《机器人》期刊 (robot.sia.cn) · Published 2024-02-04 · “一种用于可形变对象操作的动力学模型隐式求解方法”