Robotics Software Researcher (ID 1300)

Research, design, and develop advanced robot learning policies for autonomous robotic systems. Work on modern robotic intelligence architectures such as: Vision Language Action (VLA) Models World Action Models Transformer-based policy learning frameworks Build, configure, and maintain robotic simulation environments using: Gazebo Mu.

JoCo NVIDIA Isaac Sim & Isaac Lab Design and implement training pipelines using: Reinforcement Learning (RL) Imitation Learning (IL) Hybrid policy optimization techniques Train, fine-tune, and evaluate robotic policies including: Action Chunking Transformer (ACT) SmolVLA JEPA-derived robotic policies Analyse policy performance and optimize model efficiency for improved robotic task execution.

Contribute to the development of next-generation embodied AI and autonomous robotics solutions. Collaborate with cross-functional R&D teams to deploy production-ready robotic systems. Conduct experimentation, testing, and validation for real-world robotics applications.

Requirements

B.E. / B.Tech / M.E. / M.Tech / Master’s degree in: Robotics Computer Science AI Mechatronics Electronics or related fields Minimum 2+ years of experience in robotics software research or related domain. Strong understanding of: Robot learning architectures Autonomous robotic systems Embodied AI concepts Hands-on experience with: VLA Architectures World Action Models Transformer Policies Behaviour Cloning Policy Optimization Experience working with simulation platforms such as: Gazebo Mu.

JoCo NVIDIA Isaac Sim Isaac Lab Practical exposure to: Reinforcement Learning Imitation Learning ML/AI model training Strong research and experimental problem-solving mindset. Ability to convert research ideas into deployable robotic solutions. Good communication and collaboration skills. Passion for robotics, autonomous systems, and advanced AI technologies.