Kamino: GPU-based Massively Parallel Simulation of Multi-Body Systems with Challenging Topologies

2026-03-17

AI Summary

This work proposes a GPU-accelerated, high-throughput multibody dynamics simulator that natively supports complex topologies—including closed kinematic chains—and heterogeneous mechanical systems, addressing the fidelity limitations of conventional tree-structured approximations commonly used for strongly coupled constrained systems such as robots with nested loops. By formulating constrained rigid-body forward dynamics as a nonlinear complementarity problem and leveraging an efficient solver, the method enables accurate computation of contact and constraint forces. Implemented in NVIDIA Warp and integrated into the Newton framework, the simulator achieves unprecedented scalability: a single GPU can parallelize 4,096 environments of the DR Legs bipedal robot, each featuring six nested closed loops. This capability facilitates successful training of viable walking policies, demonstrating both physical realism and computational efficiency.

Technology Category

• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Machine Learning: Imitation Learning & Inverse Reinforcement Learning
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Web applications in cross-disciplinary domains and verticals such as mixed reality, smart cities, and digital health

Abstract

We present Kamino, a GPU-based physics solver for massively parallel simulations of heterogeneous highly-coupled mechanical systems. Implemented in Python using NVIDIA Warp and integrated into the Newton framework, it enables the application of data-driven methods, such as large-scale reinforcement learning, to complex robotic systems that exhibit strongly coupled kinematic and dynamic constraints such as kinematic loops. The latter are often circumvented by practitioners; approximating the system topology as a kinematic tree and incorporating explicit loop-closure constraints or so-called mimic joints. Kamino aims at alleviating this burden by natively supporting these types of coupling. This capability facilitates high-throughput parallelized simulations that capture the true nature of mechanical systems that exploit closed kinematic chains for mechanical advantage. Moreover, Kamino supports heterogeneous worlds, allowing for batched simulation of structurally diverse robots on a single GPU. At its core lies a state-of-the-art constrained optimization algorithm that computes constraint forces by solving the constrained rigid multi-body forward dynamics transcribed as a nonlinear complementarity problem. This leads to high-fidelity simulations that can resolve contact dynamics without resorting to approximate models that simplify and/or convexify the problem. We demonstrate RL policy training on DR Legs, a biped with six nested kinematic loops, generating a feasible walking policy while simulating 4096 parallel environments on a single GPU.

Problem

• multi-body systems
• kinematic loops
• parallel simulation
• constrained dynamics
• heterogeneous robots

Innovation

• GPU-based simulation
• kinematic loops
• constrained optimization
• nonlinear complementarity problem
• massively parallel reinforcement learning