Abstract: Physics-based simulation is essential for evaluating designs and control algorithms, as well as providing realism to games and movies. In robotics research, simulation has been used for evolving robotic hardware, generating datasets for supervised learning, training agents to perform long-horizon tasks via RL, and MPC. Nevertheless, there is a disconnect between the developers of robotic simulators and the roboticists that use them; namely, the simulation-builders are often unaware of the specific needs of roboticists, and roboticists are often unaware of both the capabilities and limitations of simulation. This workshop will focus on uniting these two communities to discuss one of the most challenging and important problems spanning robotics and simulation: developing simulators for deformable objects (e.g., rope, cloth, and fully-3D deformables) and leveraging them for real-world robotic manipulation. We will assess the state-of-the-art of deformable object simulation in robotics, identify unsolved problems, define benchmarks and grand challenges, and promote collaboration between roboticists and simulation researchers.