Physical Intelligence: The Next Big AI Bet
Last week, the joint venture between MEIL Group and Analog Devices marked one of India's biggest commercial bets on Physical Intelligence, an emerging field that combines artificial intelligence with robotics to enable machines to perceive, reason, and interact with the physical world.The partnership comes as global technology companies, including Tesla, NVIDIA and Figure AI, are accelerating investments in embodied AI, betting that the next phase of artificial intelligence will move beyond generating digital content to powering robots, autonomous machines and intelligent industrial systems. The partnership raises a larger question: can India build a globally competitive Physical Intelligence ecosystem, or will it remain primarily a consumer of embodied AI technologies?
Difference between Generative AI and Physical Intelligence
In contrast to generative AI, which primarily augments digital workflows by producing text, images and code, Physical Intelligence focuses on enabling machines to perceive, reason and interact with the physical world. It combines AI with robotics, computer vision, sensors, and control systems. It enables autonomous systems to perform real-world tasks in dynamic environments, and is often referred to as embodied AI.
Unlike generative AI, where errors can often be corrected by regenerating an output, Physical Intelligence operates in environments where mistakes can have immediate consequences, said Kiran Raju, CEO of autonomous airspace infrastructure Indrajaal. "If a rogue drone is approaching an airport or a border post, there's no room for the AI to hallucinate an answer," he said, adding that autonomous systems must perceive, interpret and respond to their surroundings in real time.
The ability of embodied AI to operate reliably in dynamic, real-world environments fundamentally distinguishes Physical Intelligence from generative AI. Physical Intelligence seeks to bring AI into the physical world through machines that can perceive, reason and perform real-world tasks. While global investment is increasingly flowing into robotics, autonomous manufacturing and embodied AI, India's ecosystem remains at an early stage.
India’s Emerging Ecosystem
India's robotics ecosystem is gradually expanding across manufacturing, logistics and industrial automation. Noida-based Addverb Technologies, backed by Reliance Industries, has grown into one of the country's largest robotics manufacturers, with the capability to produce up to 100,000 robots annually and exports to more than 25 countries. Indian-founded GreyOrange has meanwhile emerged as a global player in AI-powered warehouse automation, raising about USD 545 million from investors including Tiger Global Management, Mithril and Blume Ventures.
A new generation of startups is also beginning to attract investor interest. Mowito recently secured $3 million in pre-seed funding to develop AI foundation models for industrial robotic arms. At the same time, Peer Robotics, which builds collaborative autonomous mobile robots for factories and warehouses, has also raised venture funding to expand its automation solutions. Together, these companies illustrate the gradual emergence of an Indian ecosystem spanning robotics hardware, industrial automation and Physical Intelligence.
Investor’s Perspectives
Unlike software startups, Physical Intelligence ventures are evaluated not only on AI capabilities but also on their ability to build and deploy hardware at scale. Abhishek Srivastava, General Partner at Kae Capital, said investors look for founding teams that combine expertise in robotics and machine learning, alongside a clear commercial pathway for deployment. "We're willing to be patient, but we want to see a very clear first use case that's deployable, not just demonstrable," he said, adding that proprietary real-world data and successful early deployments create stronger competitive moats than software alone.
Sharing his insights on India’s opportunity, Abhishek Srivastava said, India's strengths lie in its combination of AI talent, cost-efficient research, and a large domestic market where robotics can be deployed across manufacturing, logistics, and agriculture. Companies solving problems in India's operating conditions are also likely to develop products suited for other emerging markets, he added. However, he cautioned that the country still lacks a robust hardware component ecosystem, with precision sensors, actuators and motors continuing to rely heavily on imports. "These are solvable over the next five years, but they require deliberate investment in both specialised education and domestic component manufacturing," he said.
Talent: India's Competitive Edge
India's biggest advantage in Physical Intelligence may not lie in hardware but in its deep pool of engineering and AI talent. As robotics increasingly converges with artificial intelligence, computer vision, embedded systems and control engineering, experts say the country's software expertise provides a strong foundation to participate in the next wave of intelligent machines.
According to Sandeep Kumar Sharma, Associate Professor at Thapar Institute of Engineering & Technology, student interest is already shifting beyond generative AI towards robotics and embodied intelligence, driven by advances in AI as well as growing industry demand. However, he noted that India's academic ecosystem will need to evolve to keep pace with the technology.
Traditional engineering education, Sharma said, continues to train students in silos, whereas Physical Intelligence demands interdisciplinary expertise spanning artificial intelligence, robotics, electronics, mechanical engineering and embedded systems. Universities will therefore need stronger industry partnerships, hands-on robotics laboratories and multidisciplinary curricula to build the workforce required for the sector.
From Software to Intelligent Machines
Beyond talent, experts argue that India's experience as a global software and IT services hub could provide an advantage as AI increasingly moves into the physical world.
"Generative AI answered whether machines can understand information. Physical Intelligence asks the more demanding question of whether machines can transform that understanding into reliable real-world action," said Sudhir Srivastava, Founder, Chairman and CEO of SS Innovations International.
Srivastava added, Physical Intelligence enables machines to interpret real-world conditions and act autonomously, with healthcare emerging as one of its most advanced applications. In robotic surgery, for instance, AI can combine imaging, sensor data and procedural context to assist surgeons with greater precision while keeping clinical decision-making under human supervision.
For India, the commercial opportunity extends well beyond robotics itself. As automation gains momentum across manufacturing, logistics, agriculture, healthcare and defence, demand is expected to grow for intelligent machines capable of operating safely alongside humans. With one of the world's largest industrial and services markets, the country offers a sizeable domestic testing ground for companies developing Physical Intelligence applications.
Whether India can seize that opportunity, however, will depend on more than engineering talent. Building a globally competitive Physical Intelligence ecosystem will require sustained investment in semiconductor manufacturing, precision components, robotics research, deep-tech capital and industry-academia collaboration. The MEIL–Analog partnership may be an early step in that direction, but the next phase will depend on whether the country can translate isolated successes into an integrated innovation ecosystem.
Pages:
[1]