ESTIC 2026 Session Takes Up Physical AI and Robotics

The Department of Science and Technology (DST) held a session on Autonomous Systems, Physical AI and Robotics at the Emerging Science, Technology and Innovation Conclave (ESTIC) 2026 at Bharat Mandapam, New Delhi. Experts at the session weighed how machines that act on their own, physical artificial intelligence and robotics will reshape research and the making of new technology in the years just ahead. The conclave itself ran from 27 to 29 October 2026 and drew people from science, research, industry and the start-up world.

What the DST Secretary said

Prof. Umesh V. Waghmare, Secretary of the Department of Science and Technology, opened the session. He argued that industry, universities, researchers, start-ups and the users of technology have to work as one if autonomous systems, physical AI and robotics are to move ahead, and if ideas from the laboratory are to turn into answers that help society. That meeting of minds, he said, matters most at the point where a finding has to become something usable.

The cyber-physical systems mission

Prof. Waghmare drew attention to DST's National Mission on Interdisciplinary Cyber-Physical Systems (NM-ICPS), under which technology innovation hubs have been set up at 25 leading academic institutions across the country. The hubs put academia in touch with industry and start-ups so that laboratory ideas reach the market. Phase I of the mission is to close in December 2027, and he said the views gathered at the session would feed into the plan for the second phase.

Physical AI is a systems problem

Prof. Abhay Karandikar, Member of NITI Aayog, told the session that physical AI is not only a question of artificial intelligence but equally one of systems engineering. An intelligent physical machine works only when many parts hold together: dependable sensors, efficient actuators, computing that keeps up in real time, sturdy control algorithms, communication with little delay and hardware that is sparing with energy. He drew a sharp line between such systems and conversational AI tools, because a mistake by a robot on a factory floor or by a self-driving vehicle has an immediate physical cost. Reliability, verification, safety, security and trustworthiness must therefore stay at the centre of work in this field.

A Physical AI Mission on the table

Prof. Karandikar said India can build autonomous systems on strengths it already has in systems engineering, electronics, manufacturing, embedded systems, design, telecommunications and software. He added that consultations with stakeholders have examined a proposal for a dedicated Physical AI Mission, and suggested that the next phase of NM-ICPS could be the place to take up this field. The session, numbered six at the conclave, had a keynote address, invited talks and a panel discussion, with speakers from Addverb Technologies, IIT Hyderabad, IIT Delhi, IIT Indore and robotics firms.