en.Wedoany.com Reported - Jaiveer Singh, a robotics software engineer at Nvidia, is leading the development of Isaac ROS, the company's open-source robotics software stack built on ROS 2. He leads a team within Nvidia that develops software for autonomous mobile robots, manipulation systems, and humanoid robots using the company's hardware and software tools. The platform combines CUDA-based libraries and AI models with ROS 2. Describing the work, Singh emphasized that its core is infrastructure, including the compute boards inside the machine, software for processing robot camera data, and the engineering required to advance the system from prototype to repeatable deployment.
Isaac ROS began as an internal experiment during Singh's internship on the robotics team. After completing studies in electrical engineering, computer science, and business at the University of California, Berkeley, he joined Nvidia full-time and continued working on the project that later became the company's robotics software stack. A key part of the approach is open-source release, designed to allow developers to inspect, adapt, and extend the software, rather than relying on closed systems that may not meet long-term needs. Singh stated that the initial question for the project was whether robotics software built around the Nvidia Jetson platform and Nvidia CUDA libraries could create value as an open-source release. The answer became clear after developers leveraged more processing power from graphics hardware, and he believes developers always want to unlock the full performance of GPUs.
Isaac ROS adopts a modular design, supporting multiple robotics development categories such as manipulation, mobility, and humanoids. The software includes packages for perception, object detection, mapping, collision detection, and motion planning, and can run on workstations, DGX Spark systems, and Jetson edge devices. Unlike the original Isaac SDK, Isaac ROS is designed as a fully modular system. Singh used a Lego brick analogy, stating that developers can combine Nvidia's packages with internally created or existing ROS code from the broader robotics community.
The open-source model has become a core strategy for Nvidia in promoting Isaac ROS. Facing an industry with rapidly changing product demands and technical standards, developers and startups need assurance that the software will remain usable and adaptable years into a project. In robotics, companies typically build products over long development cycles and need to integrate software with sensors, actuators, and safety systems. The ability to inspect and modify code, as well as contribute fixes, is as important as having the original tools. Nvidia is adapting Isaac ROS to a market focused on humanoid robots and AI agents, making the software more suitable for developers building these systems and meeting the perception, planning, and control needs of full-stack software across diverse hardware environments.
Singh noted that Nvidia's early investment in robotics was one of the reasons he was attracted to the company. He believes that before the current surge of interest in physical AI, Nvidia had already been deeply involved in the field and committed to solving problems. For Singh, the significance of open source in robotics lies in spreading confidence and responsibility, enabling fixes and improvements from one company to benefit the broader industry facing similar technical challenges.
