DLO-Lab: Benchmarking Deformable Linear Object Manipulations with Differentiable Physics
2026-06-02 • Robotics
Robotics
AI summaryⓘ
The authors created a new computer simulator that helps robots learn to handle flexible objects like ropes and cables, which are hard to work with because they can bend and stretch in many ways. Their simulator can imitate different material behaviors and interactions, making it more useful than previous tools. They also designed a set of test tasks showing common difficulties in manipulating these objects. To solve these tasks, the authors developed a special robot control system that picks smart points to grab and breaks big jobs into smaller steps. They tested different learning methods and showed that their approach helps robots better learn and apply skills both in simulation and real life.
deformable linear objectsdifferentiable simulatormaterial propertiesrobot manipulationgrasp planninglong-horizon taskspolicy learningsim-to-real transferelasticitybending plasticity
Authors
Junyi Cao, Yian Wang, Ziyan Xiong, Chunru Lin, Zhehuan Chen, Chuang Gan
Abstract
We address the challenge of enabling robots to manipulate deformable linear objects (DLOs), such as ropes, cables, and rubber bands. Prior work has primarily focused on narrow, task-specific problems, often relying on real-world demonstrations or handcrafted heuristics. Such approaches, however, struggle to scale to the wide variety of materials and tasks encountered in practice, and collecting sufficiently diverse real-world data is often impractical. Additionally, existing simulation environments offer limited support for the broad spectrum of material behaviors necessary for generalizable DLO manipulation. To overcome these limitations, we introduce a differentiable simulator explicitly designed for versatile DLO manipulation. Our simulator models a wide range of material properties-including (in)extensibility, elasticity, bending plasticity, and complex interactions with other objects-providing a robust foundation for learning and evaluating manipulation skills. Building on this simulator, we propose a benchmark suite of representative tasks that highlight the unique challenges of DLO manipulation. The successful execution of these tasks is often hindered by the topological complexity and grasp sensitivity inherent to DLOs. Therefore, we introduce a specialized DLO agent that explicitly manages these challenges by proposing strategic grasping points and decomposing long-horizon tasks to maximize control authority. Finally, we evaluate various policy-learning algorithms using our framework, alongside sim-to-real transfer experiments, demonstrating our platform's potential to advance DLO manipulation.