[Project] newton_ros, ros2 bridge for the Newton simulator

Developer vybhav_ilindra has released newton_ros, a significant open-source bridge that directly connects the Newton physics simulator to the ROS2 (Robot Operating System 2) control framework. Newton is a GPU-accelerated physics engine built on NVIDIA's Warp framework and is maintained by the Linux Foundation with contributions from NVIDIA, Google DeepMind, and Disney Research. The key innovation is that newton_ros is a standalone RCLPy package, meaning developers can leverage high-fidelity, GPU-powered physics for robotics simulation without the overhead or licensing constraints of NVIDIA's Omniverse platform. It utilizes the topic_based_ros2_control by Picknick Robotics and supports a wide array of critical sensors including RGB/depth/segmentation cameras, 3D LiDAR, IMUs, and contact sensors.

The bridge dramatically lowers the hardware barrier for advanced robotics simulation. While it supports CUDA-enabled GPUs for optimal performance, it can also run on CPU—albeit slowly. In a practical test, vybhav_ilindra achieved a smooth 30 FPS in a scene with an active camera, 3D LiDAR, and viewer on a modest RTX 3050 Mobile laptop GPU. This performance makes sophisticated simulation accessible to a broader range of researchers, students, and indie developers. The project reuses existing simulation interfaces, promoting interoperability within the ROS ecosystem and offering a compelling, vendor-agnostic alternative to proprietary simulation stacks like NVIDIA Isaac Sim for prototyping and testing ROS2-based robotic systems.