After transforming digital work through chatbots and content-generation tools, artificial intelligence is now moving into the physical world, pushing investors and companies to bet on autonomous machines that can drive vehicles, operate warehouses, inspect factories and make real-time decisions on the ground.
The shift, increasingly referred to as physical AI, is drawing attention globally after comments from Jensen Huang and strategic moves by Tesla signalled that the next phase of AI may lie beyond software. Physical AI refers to AI systems embedded in machines such as robots, drones, cameras and autonomous vehicles that can sense, reason and act in the real world rather than simply generate text, images or code on screens.
Nvidia CEO Jensen Huang recently said the company was pivoting towards physical AI because “unless the computer can understand the physical world, there is no chance of enhancing it”. Tesla, meanwhile, is repositioning itself as a robotics and AI company, with CEO Elon Musk betting heavily on humanoid robots under its Optimus programme.
Macro Shifts and Platform Evolutions
The global pivot is beginning to influence Indian startups and venture capital firms, which are increasingly viewing physical AI as the next major deep-tech opportunity after generative AI. Unlike the first wave of AI applications centred on digital productivity, physical AI is being pitched as a way to improve operational efficiency in sectors such as manufacturing, logistics, mobility, agriculture and industrial safety.
“Physical AI is plausibly the next platform shift after language. We expect to be active here over the coming quarters,” Sonal Saldanha, principal at 3one4 Capital, told FE. The firm is evaluating opportunities across manufacturing, logistics, healthcare and agriculture, particularly where businesses are willing to pay for measurable improvements rather than wait for perfect automation systems.
The emerging investor interest reflects a broader shift in the economics of AI. While generative AI demonstrated how machines could automate knowledge work, companies are now looking for systems that can reduce wastage, improve safety, increase factory precision and automate repetitive physical tasks.
Experts say three developments are driving the transition. AI models have become significantly more efficient, sensors and edge-computing hardware have become cheaper, and businesses have become more willing to experiment with AI after the mainstream adoption of generative AI tools.
“Gen AI made businesses move faster, experiment more aggressively, and start asking a much bigger question: where can AI create measurable operational value? That shift is now pushing focus toward Physical AI,” said Teja Gudena, executive vice-president, engineering, at Netradyne.
Netradyne’s Driver•i platform, deployed across more than 500,000 commercial vehicles globally, uses AI-powered cameras and sensors to analyse driver behaviour and road conditions in real time. Similar systems are increasingly being adopted in warehouses, industrial facilities and manufacturing lines.
Indian Testing Ground
The trend is also beginning to shape startup funding activity in India. Last month, IAN Alpha Fund led a Rs 22 crore seed funding round in RoshAi, which develops retrofit-based driverless solutions for industrial vehicles. This week alone, semiconductor startup HrdWyr, construction robotics startup A Flo Mobility and robotics firm Alphadroid announced fresh fundraises.
Large industrial groups are also beginning to deploy such systems. Reliance Industries has deployed Nvidia Omniverse technology at Jamnagar, while Tata Steel has used computer vision systems to reduce defects at its Jamshedpur plant.
Investors say the immediate opportunity may not lie in humanoid robots, which continue to face uncertain commercial viability, but in narrower industrial applications where companies can quickly see cost savings and productivity gains. As AI moves from chat windows to factory floors, India’s manufacturing and logistics sectors are increasingly emerging as the next testing ground for the technology.