Inner Sky Labs, the Mumbai company behind the Miko companion robot, has come out of stealth with a stack of foundation models it says lets machines sense their surroundings, decide what to do and act safely around people. The company announced the platform. The stack has three parts, according to the company. Its perception models include the Drishti series for vision, the Shruti series for audio, such as spoken instructions or the sound of something falling, and the Sparsh models for touch, which it says let a robotic hand tell a grip on an egg from a grip on a wrench. Its action models, Kriya, Prana and Karma, are meant to plan and carry out physical movement such as reaching, gripping, walking and flying, and to predict what happens next, for example whether a box will slip or a person nearby will step forward.

The third part is Kavach, Sanskrit for “shield”, which Inner Sky Labs describes as a safety layer that checks every action a machine is about to take and can stop it regardless of what the rest of the system has decided. In the company’s example, before a robotic arm swings across a factory floor, Kavach checks whether the space is clear of people, whether the planned speed is safe for that area and whether the action stays inside the zone the arm is allowed to work in. If any answer is no, “the arm does not move,” the company says, and the action is blocked before it starts, not slowed or flagged afterwards. It also says any customer or authority can write its own safety rules, in its own legal language, and have a machine obey them.

Inner Sky Labs ties the safety approach to Miko, which it says has been used by children in more than 140 countries over ten years. It says an independent evaluation by the cyber security practice of BDO found that Miko’s safety system caught every unsafe prompt across seven public AI-safety benchmarks, ahead of four widely used commercial AI models, and ranked first at flagging unsafe AI responses. The company points to www.innerskylabs.ai for the methodology, which was not part of the release. It says the full stack already runs in factories, classrooms and consumer products.

“Every robot company will eventually have to answer one question: why should anyone trust this machine near a person,” said Sneh Vaswani, co-founder and chief executive of Inner Sky Labs. “For ten years, our answer was Miko — a robot trusted enough to be left with a child. Today we’re answering that same question for every kind of machine, with technology built, tested and proven out of India.”

Keshav R. Murugesh, an early investor and advisor to the company and former Chairperson of NASSCOM, said: “I’ve advised Sneh and the team since 2016, when they first shared this multi-decade vision with me—long before ‘Physical AI’ was even a category. Today, Physical AI is being called a $50 trillion opportunity. I am proud to see this built in India for the world. Miko was the proof point. This is the scale-up.”

The platform is available to qualified partners by invitation, with cloud, on-premise and on-device deployment, including offline, air-gapped environments. Each model, including Kavach, can be used on its own. The company says it has raised over $80 million from investors including Chiratae Ventures, IvyCap Ventures, YourNest Venture Capital and 360 ONE. Pricing and customer names were not disclosed.