Raisim

Raisim is a physics-based robotics simulation framework that provides real-time capable simulation of rigid-body systems with contacts. It is used to develop and test control, planning, and learning algorithms for robots in simulation before deployment on real hardware.

The engine supports modeling of rigid bodies, joints, actuators, and sensor-like outputs, along with stable contact resolution and friction models. It offers both C++ and Python APIs, enabling fast prototyping and easy integration with machine learning workflows. Raisim emphasizes efficiency and accuracy, making it suitable for iterative experimentation and real-time interaction.

Raisim is designed to be lightweight and performant, with a focus on fast simulation steps and low overhead. This makes it practical for running large numbers of trials, simulations in closed-loop with controllers, and experiments that require responsive feedback.

In research and education, Raisim is used for robotics control development, reinforcement learning experiments, planning and perception studies, and sim-to-real investigations. Its flexible API and modular design allow users to build and simulate diverse robotic systems, environments, and tasks, contributing to a broad range of applications in academia and industry.