AuraML has unveiled AuraSim, a multimodal world simulation platform designed to generate physics-ready environments for robotics training and deployment. Built using NVIDIA AI infrastructure, the platform enables developers to create high-fidelity digital simulation environments from inputs such as text descriptions, floor plans, videos, depth maps, and point clouds.
Positioned as a foundational infrastructure layer for robotics and logistics applications, AuraSim aims to accelerate the development and deployment of physical AI systems by enabling large-scale simulation before real-world implementation. Alongside the launch, AuraML also announced the Global Physical AI & Robotics Cohort, a training initiative developed in collaboration with NVIDIA Inception to build skills in robotics simulation and physical AI development.
Bridging digital and physical environments
Training autonomous systems requires scalable simulation environments that accurately replicate physical conditions. AuraML’s platform converts multimodal inputs into “sim-ready” digital twins designed to reflect real-world physics, enabling robotics developers to test scenarios, optimise workflows, and reduce deployment risks.
Key capabilities of the AuraML World Simulation Model include:
- Multimodal environment generation: Creation of complex 3D environments from floor plans, text prompts, or video inputs.
- Physics-ready simulations: high-fidelity outputs with accurate collision boundaries and physical properties.
- Agentic automation: AI-driven pipelines capable of generating large-scale simulation scenarios autonomously.
- Rapid iteration: Developers can stress-test edge cases and workflows through prompt-based simulation design.
Supporting robotics and industrial automation
The platform is designed to simulate complex industrial environments, enabling organisations to test robotics workflows before deploying physical systems. In factory automation scenarios, AuraSim allows simultaneous simulation of humanoid robots, autonomous mobile robots (AMRs), and stationary robotic arms operating within a unified virtual environment — helping ensure synchronisation, safety, and operational efficiency.
Powered by NVIDIA Technologies
AuraSim is built on NVIDIA Omniverse and NVIDIA Isaac Sim, leveraging NVIDIA Cosmos world foundation models and accelerated computing capabilities. The integration enables photorealistic simulation environments and seamless interaction between digital twins and real-world robotics systems.
To further accelerate adoption, AuraML and NVIDIA are launching an eight-week Global Physical AI & Robotics Cohort, aimed at robotics engineers, AI researchers, and developers. The programme will focus on building, simulating, and deploying real-world physical AI solutions through mentorship-driven training.
Ayush Sharma, co-founder of AuraML, said the launch represents a major milestone for physical AI development, enabling robotics companies to move from concept to deployment faster through integrated simulation workflows. Tobias Halloran, Director of EMEAI Startups and Venture Capital at NVIDIA, highlighted India’s growing AI startup ecosystem and NVIDIA’s efforts to support founders through accelerated computing and innovation programmes such as NVIDIA Inception.
The announcement comes amid growing momentum around the India AI Impact Summit 2026, where discussions focused on sovereign AI infrastructure, robotics, and next-generation innovation ecosystems.