Inner Sky Labs Launches Physical AI Platform, Seeks ₹300 Cr Funding

Inner Sky Labs is also closing a ₹300 crore funding round involving existing backers YourNest Venture Capital and Keshav R. Murugesh, along with new global investors. After spending more than a decade developing robotics technology, the company is now taking its work beyond consumer products and into factories, warehouses, drones and other machines that operate around people.

From Understanding Instructions To Taking Action Most AI tools people use today work with information. They answer questions, analyze documents or generate content. Physical AI has a different job.

It must understand what is happening around a machine and help it respond. Think of a robot moving boxes inside a warehouse. It needs to recognize the box, know how firmly to grip it and check whether someone is standing in its path.

It must also adjust if something changes while it is working. Inner Sky Labs has developed several AI models to handle these tasks. Its Drishti models help machines interpret visual information, while Shruti processes sounds, including spoken instructions and noises in the surrounding environment.

Another set of models, called Sparsh, focuses on touch. That last capability is particularly useful for machines handling different objects. Picking up an egg, for instance, requires a very different grip from lifting a wrench.

The company has also introduced three action models called Kriya, Prana and Karma. These are designed to help machines plan movements such as reaching, gripping, walking and flying. The models also attempt to predict what could happen next.

A robot moving a box, for example, would need to account for whether the box might slip or whether a person could step into its path. Bringing these capabilities together is what Inner Sky Labs hopes will make machines more useful in everyday working environments. Kavach Checks Before A Machine Moves Of all the technologies announced, Kavach addresses perhaps the most difficult question in physical AI: how do you stop a machine from taking an unsafe action?

Inner Sky Labs has developed Kavach as a separate safety system that checks a machine's planned movements before allowing them to happen. Consider a robotic arm working on a factory floor. The machine may decide to move an object from one place to another.

Before it acts, Kavach checks whether the movement is within the permitted area, whether the speed is safe and whether someone is standing nearby. If the planned movement breaks a safety rule, Kavach is designed to block it. The idea is to give machines a separate layer of control rather than leaving every decision to the AI system responsible for performing the task.

The company says businesses and authorities can also define their own safety rules, allowing machines to follow requirements specific to a factory, industry or country. That could be useful for organizations that need tighter control over automated equipment, particularly where machines work close to employees. But the real question is how reliably such a system can detect danger when conditions change unexpectedly.

Inner Sky Labs has not shared detailed independent industrial testing results for Kavach in its announcement.

About AI Safety Inner Sky

Labs is entering this market with experience in consumer robotics. Its Miko companion robot has spent the past decade interacting with children and is now present in more than 140 countries, according to the company. Building a robot for children brought its own safety challenges.

Miko needed controls over the kind of responses it could give and the conversations it could have with young users. Inner Sky Labs says that experience helped shape the thinking behind Kavach. The company also pointed to an independent evaluation by BDO's cybersecurity practice, in which Miko's safety system detected every unsafe prompt across seven public AI safety benchmarks.

It also ranked first in identifying unsafe AI responses when compared with four widely used commercial AI models. Those findings relate to AI response safety, however, rather than the physical movement of industrial robots. The two involve different risks, and the performance of one does not automatically establish the reliability of the other.

For Inner Sky Labs, the challenge now is to take what it learned from Miko and apply it to machines operating in much less predictable environments.

"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." The move also changes the kind of customers Inner Sky Labs will need to convince. A consumer choosing a companion robot and a factory operator deploying automated machinery have very different expectations of safety and reliability. ₹300 Crore Funding For The Next Phase Alongside the technology launch, Inner Sky Labs is working to close a ₹300 crore funding round.

Existing investors YourNest Venture Capital and Keshav R. Murugesh are involved, along with new global investors whose names have not been disclosed. The company has previously raised more than $80 million from investors including Chiratae Ventures, IvyCap Ventures, YourNest Venture Capital and 360 ONE.

The fresh funding comes as Inner Sky Labs prepares to offer its technology to a wider range of businesses. Keshav R. Murugesh, an early investor and advisor to the company and former Chairperson of NASSCOM, has been associated with the team since 2016.

“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,” said Keshav R. Murugesh, early investor, advisor to Inner Sky Labs, and former Chairperson of NASSCOM. “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.” While Murugesh referred to the potential size of the physical AI opportunity, the company's immediate focus will be on turning its technology into products that industrial customers can deploy and maintain.

Inner Sky Labs has not disclosed its valuation, the full investor list for the proposed round or how the fresh capital will be allocated. Bringing The Technology To Businesses Inner Sky Labs says its physical AI platform is available to qualified partners by invitation. Businesses can run the models through the cloud, on their own infrastructure or directly on devices.

The platform also supports offline and air-gapped environments, which can be useful for facilities that cannot depend on external network connections. Companies do not have to adopt the entire platform either. Inner Sky Labs is making its perception models, action models and Kavach safety system available separately.

This gives businesses the option to use specific capabilities depending on their existing machines and operational needs. For example, a company may need better visual recognition for a robotic system without replacing its movement controls. Another may be interested primarily in adding a separate safety check.

The company says its technology is already being used in factories, classrooms and consumer products in millions of homes. However, it has not disclosed how many industrial customers are currently using the new platform or provided details of specific factory deployments. Those numbers would help show how far the technology has moved beyond development and into regular commercial use.

The Real Test Will Be On The Factory Floor Physical AI is an interesting direction for robotics, but making it work reliably outside controlled settings is not easy. Factories and warehouses are busy places. People move around, equipment behaves differently over time and unexpected situations are part of everyday operations.

A machine needs to handle those situations without putting people or equipment at risk. Inner Sky Labs is trying to solve part of that problem by separating a machine's ability to act from the system that decides whether the action should be allowed. The approach gives businesses a way to define safety rules, but its success will depend on how well those rules work in real environments and how easily the technology fits into existing machinery.

As more AI systems begin controlling machines rather than simply responding to instructions, those questions will become part of everyday technology decisions. For Inner Sky Labs, answering them on actual factory floors will matter more than what its models can demonstrate in a lab.