DLO-Lab: Benchmarking Deformable Linear Object Manipulations with Differentiable Physics
For robotics researchers working on deformable object manipulation, this provides a more versatile simulation and benchmarking platform, though it is an incremental step combining existing differentiable physics with DLO-specific design.
The paper introduces a differentiable simulator for deformable linear objects (DLOs) that models a wide range of material properties, along with a benchmark suite of tasks and a specialized DLO agent. The framework enables evaluation of policy-learning algorithms and shows promise for sim-to-real transfer.
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.