Amazon Web Services (AWS) today launched an open-source Physical AI Toolchain designed to help roboticists move from collecting data to training, simulating, validating and deploying AI models on their robots.
The toolchain brings together AWS services and NVIDIA’s Physical AI software stack into a single development workflow. The goal is to address one of the biggest challenges facing companies developing AI-powered robots: connecting the many pieces required to turn a trained model into a system that can reliably operate in the real world.
“We’re trying to make this easy button, if you will, so that people can actually focus on the problem that they’re trying to solve rather than having to worry about all of the infrastructure,” Sri Elaprolu, director of Frontier AI Science and Engineering at AWS, told The Robot Report.
The Physical AI Toolchain is not a direct replacement for RoboMaker, a cloud-based robotics simulation platform that was shut down in 2025. Elaprolu described RoboMaker as one of the components of AWS’s robotics stack, while the new toolchain brings together a broader set of development technologies.
The toolchain combines five pieces of the AI development process: synthetic data generation, model training, simulation and validation, edge deployment, and continuous improvement. A robotics company could collect demonstrations of a robot performing a task, generate additional training scenarios synthetically, train or fine-tune a model, and test that model in simulation before deploying it to a physical robot.
AWS uses Amazon SageMaker for model training and AWS IoT Greengrass for distributing models to edge devices.
Source link







