Robots have been used in factories for years, but most are built for one fixed task in one controlled space. Tensrai wants to make industrial machines more flexible, capable, and easier to deploy.

The Hyderabad-based startup has raised Rs. 3.4 crore in a Pre-Seed round to develop what it calls physical AI – technology that allows robots and industrial vehicles to understand their surroundings and carry out real-world work with less human control.

What Tensrai is building

Tensrai was founded in 2022 by Sai Nikhil Rao Gona, Venkata Sai Charan Chalamalasetty, Harshith Uppalapati, and Vivek Datta.

The company is developing an autonomy platform for robots, industrial vehicles, and machines. In simple words, it wants to give machines the ability to move safely, understand their environment, and complete useful physical tasks.

For example, a robot in a warehouse may need to move boxes from one area to another. In a factory, it could transport components between workstations. In an airport or hospital, it could support cleaning or material movement.

Tensrai’s platform is designed to work with new robots as well as existing industrial equipment. This retrofit approach is important because businesses may not want to replace every forklift, trolley, or machine they already own. Instead, they may be able to add autonomous features to suitable equipment.

Where the Pre-Seed money will go

As per Newzchain, the Pre-Seed round was backed by India Accelerator, Finvolve, Hyderabad Angels, and other angel investors. Tensrai plans to use the capital for technology development, engineering, deployment capabilities, and expansion in India and overseas markets.

The funding should help the startup improve the hardware and software needed to make its robots work in real environments. A warehouse floor is not always smooth, quiet, or predictable. There can be people walking around, moving vehicles, uneven surfaces, and changing layouts.

This is where physical AI differs from a regular software tool. A chatbot works with information on a screen. A physical AI system must understand the world around it and take action safely. Tensrai says its technology uses cameras and software to help machines navigate, identify surroundings, and perform tasks with limited human intervention.

The company also plans to earn money through robot sales, retrofitting existing machines, customized projects, and Robot-as-a-Service. Under the Robot-as-a-Service model, a customer pays for the use of a robot instead of buying it outright. That could make automation more affordable for businesses that do not want a large upfront investment.

The problem Tensrai wants to solve

Factories, warehouses, and large commercial sites often face shortages of skilled operators for repetitive or physically demanding work. Manual material movement can also be slow, especially when operations run around the clock.

Tensrai’s aim is not simply to replace workers. The stronger use case is helping people avoid repetitive, heavy, or potentially hazardous work while allowing teams to focus on supervision, maintenance, and higher-value tasks.

The startup will compete with warehouse-automation firms, industrial robotics companies, and makers of autonomous mobile robots. Its key challenge will be proving that its systems remain reliable across different sites and work conditions.

A robot that works well in a controlled demonstration must also work safely in a busy factory or uneven outdoor industrial yard. That is where practical deployment matters more than ambitious claims.

Mowito has raised $3 million in pre-seed funding to build AI software for industrial robots. The startup is working on a simple but powerful idea – factory robots should learn a task by watching a human do it, instead of needing fresh code every time the production line changes.

That may sound like a small shift, but in manufacturing it can save time, money, and engineering effort. Today, many factory robots are useful but not very flexible. They can repeat tasks well, but when a product design changes or a new process is added, engineers often need to reprogram them.

Mowito wants to reduce that burden. Founded in 2024 by Puru Rastogi, Adityanag Nagesh, and Safar V, the startup is building AI foundation models for industrial robotic arms. It currently works across Bengaluru and Detroit, with customers in automotive and electronics manufacturing.

Why this funding matters

The $3 million round was led by Version One Ventures. All In Capital, Unisol, and iSeed also joined the round. Angel investors included AI researcher Soumith Chintala, along with founders from Foundry Robotics, Coformer.ai, and Better Capital.

For an early-stage startup, pre-seed money is usually used to build the core product, hire the right team, test with customers, and prepare for wider market entry. In Mowito’s case, the funding will support its expansion in the United States, grow its engineering and go-to-market teams, and help it deepen deployments with manufacturing customers.

This is important because industrial robotics is not a quick app business. A factory robot has to work safely, repeatably, and with very little room for error. If a robot fails on a production line, it can delay output and hurt quality. So Mowito has to build slowly, test carefully, and prove that its software works in real factory conditions.

What Mowito is trying to solve

Factories already use robotic arms for welding, lifting, picking, placing, inspection, and assembly. These robots are strong, accurate, and tireless. But they are often difficult to adapt.

For example, an electronics manufacturer may change the design of a device part. A car parts supplier may add a new component to a line. A factory may shift from one product batch to another. Each change can create extra work for engineers because the robot may need fresh instructions.

Mowito’s approach is different. The company is building software that can run on standard industrial robotic arms without changing the hardware. Instead of writing detailed code for every task, a human can demonstrate the task, and the robot can learn from that demonstration.

In simple words, it is like teaching by showing.

Why “teaching by demonstration” is useful

Imagine a worker showing a robotic arm how to pick up a component, rotate it, and place it into a fixture. In the old way, an engineer may need to write or adjust code so the robot understands every movement.

With demonstration-based learning, the robot can observe the task and repeat it. The goal is to make robots easier to train when production needs change.

This can be especially useful for small and mid-sized manufacturers. Large factories may have big automation teams, but smaller plants often cannot afford long programming cycles. If robot training becomes simpler, more factories can use automation without needing a large robotics team.

That is where Mowito’s purpose becomes clear. It wants to make industrial automation more flexible and easier to use.

How this can support automotive and electronics manufacturing

Mowito is already focused on automotive and electronics customers, which makes sense. Both sectors have many repeatable tasks, but they also change quickly.

In automotive manufacturing, robots may work on parts, sub-assemblies, welding, inspection, or handling. Vehicle platforms keep changing, and suppliers often manage different parts for different customers.

In electronics, parts are smaller, designs change fast, and production lines may need frequent updates. A robot that can learn faster can help teams reduce downtime.

For example, if a factory is making phone components and the part shape changes, the robot should not become useless until an expert rewrites its program. If it can learn from a fresh demonstration, the production team can move faster.

This is the kind of practical problem Mowito is trying to solve.

The bigger purpose behind the startup

Mowito’s larger aim is not just to make robots smarter. It is trying to make factories less dependent on slow robot programming.

That purpose matters because manufacturing is under pressure. Companies want faster production, better quality, lower costs, and more flexible lines. At the same time, skilled automation engineers are not always easy to find.

If Mowito can make robotic arms easier to train, factories can respond faster to new orders and product changes. Workers can spend less time waiting for reprogramming and more time improving production.

There is also a wider India angle. If Indian startups can build serious industrial AI from India while also serving markets like the US, it strengthens the country’s position in deep-tech and manufacturing software.

What “physical AI” means

Mowito is often described as a physical AI startup. This simply means its AI is built for the real world, not only for text, images, or chat.

A chatbot gives answers on a screen. A physical AI system helps a machine act in the real world. It may use camera input, sensor data, movement data, and human demonstrations to decide what a robot should do next.

This is harder than regular software because the real world is messy. A part may be slightly tilted. Lighting may change. A tool may be in a different position. A human may stand nearby. The robot has to understand the situation and still act safely.

That is why the opportunity is large, but the challenge is also serious.

Competitors and market pressure

Mowito is entering a fast-growing area where many companies are trying to make robots easier to use. Global names in nearby areas include Physical Intelligence, Covariant, Mech-Mind Robotics, Mujin, Micropsi Industries, and Standard Bots. Some focus on robot intelligence, some on vision systems, some on warehouse automation, and some on easier robot programming.

Mowito’s edge will depend on how well its software works on standard robotic arms, how quickly customers can train robots, and how safely the system performs in real factory settings.

The market will not reward fancy demos alone. Manufacturers care about uptime, accuracy, speed, safety, and return on investment. If Mowito can show clear savings in time and cost, it can build trust.

What to watch next

The next phase will be important for Mowito. The startup has to expand in the US, grow its team, and keep proving its product with real customers.

A few things will matter most. Can the software handle different robot brands? Can it work across different factory tasks? Can non-experts train robots without trouble? Can it reduce downtime enough for customers to pay for it again and again?

If the answer is yes, Mowito could become part of a larger shift in manufacturing, where robots are no longer hard-coded for every small change but trained more like human workers.

Conclusion with key takeaways

Mowito’s $3 million pre-seed funding is an important signal for industrial AI and robotics. The startup is trying to solve a real factory problem – robots are powerful, but programming them is still too slow and costly.

By helping robots learn from human demonstrations, Mowito can make automation easier for manufacturers in automotive, electronics, and other production-heavy sectors.

Key takeaways

• Mowito has raised $3 million in pre-seed funding.
• The round was led by Version One Ventures.
• The startup was founded in 2024 by Puru Rastogi, Adityanag Nagesh, and Safar V.
• Mowito builds AI software for standard industrial robotic arms.
• Its technology helps robots learn tasks by watching human demonstrations.
• The company works across Bengaluru and Detroit, with customers in automotive and electronics manufacturing.

Facts Input- ET, ArXiv