How Ottonomy Is Rewriting Last-Mile Delivery

Vijay brought 18 years of robotics industry experience to the table when he and his co-founders, Hardik Sharma, Pradyot Korupolu, and Ashish Gupta, began working on their idea in 2020.

The timing was both a challenge and an opportunity. With large parts of the world under lockdown, the team built their first robot in a guest room in India, ran a test in the basement, and booked their first pilots with e-commerce companies.

Their observation of the market was sharp. Warehouses, manufacturing, and autonomous vehicles had attracted significant robotics investment. But hyperlocal delivery and logistics, the last hundred metres of a supply chain, remained largely untapped.

"Back in 2020, only a few delivery robotics companies were focused on food delivery. Their scope was limited by the size and capabilities of the robots, as most relied on GPS and similar tools for localisation," Vijay says.

Early pilots followed with companies like Walmart and several airports. But Vijay quickly realised that the unit economics did not work for food delivery. The company pivoted its focus toward healthcare and industrial deployments, where the value of reliable, autonomous delivery was far more compelling.

Ottonomy is based in the US but manufactures everything from scratch in India, working entirely with homegrown founders. About 40% of the supply chain is domestic, with critical components like LiDAR sensors and semiconductors imported. The company has raised $7.8 Mn from investors including pi Ventures, CoreNest, Connetic Ventures, and ADR Ventures.

Most delivery robots navigate using GPS or pre-mapped routes. They know where they are. What they often do not know is what kind of environment they are in and how to behave within it.

Ottonomy's approach is different. Its robots use what the company calls Contextual AI: pre-trained models that help the robot understand its surroundings, whether it is a hospital corridor, a shopping mall, or a public sidewalk. Once the context is identified, a reinforcement learning pipeline governs real-time behaviour, deciding how the robot should move, yield, prioritise routes, or avoid obstacles based on continuous feedback loops.

In practice, this means the robot learns whether to yield right-of-way to a wheelchair, how to navigate a crowded airport terminal, or when to slow down near an elevator door. It is not just about reaching a destination. It is about behaving appropriately along the way.

The robots are also built to operate in demanding weather conditions. Ottonomy recently deployed a fleet at a chemical facility in Finland, where temperatures dropped to minus 18 degrees Celsius and the robots moved goods between buildings through snow.

"The robots work absolutely fine. They can take different weather conditions, from cold, rain, and even heat," Vijay says.

Ottonomy operates with two primary robot models: Ottobot 2.0, designed for industrial environments, and Ottobot 3.0, built with a narrower form factor to navigate tighter spaces like hospital elevators and corridors.

Rather than building a separate robot for every use case, the company uses customisable compartment modules mounted on top of the base platform. With 6 to 8 compartment configurations, the same robot can be adapted for multi-order last-mile deliveries of up to 8 to 10 packages in a single trip, secure medical transport including blood samples, chemotherapy kits, and vaccines, warehouse and industrial material movement, and high-value payload delivery.

This modular approach keeps manufacturing costs manageable and makes deployments faster to set up and reconfigure.

Ottonomy has filed 29 patents and had 24 granted, covering various aspects of robotics, autonomy, and system design.

Ottonomy also runs Ottumn.ai, a fleet management and orchestration platform that works not just with its own robots but also with drones, robotic arms, smart mailboxes, elevators, and access doors.

The platform allows enterprises to onboard different types of robots, integrate their APIs, and coordinate how they work together rather than in silos. It is, in effect, the operating system for a multi-robot, multi-device deployment.

Subscription fees for Ottumn.ai range from $100 to $800 per month per system, adding a recurring software revenue stream alongside the hardware business.

Rather than selling robots outright, Ottonomy operates on a Robots-as-a-Service model. Enterprises lease robots through monthly subscriptions priced at approximately $999 per robot per month on contracts ranging from one to five years.

Before committing to a long-term contract, customers run a paid pilot lasting one to three months. The model is available across the US, UK, Europe, Australia, and India.

While the US remains Ottonomy's largest market, the company is actively building its India presence. It is running a pilot at the Hyderabad airport and has recently partnered with Skye Air Mobility and Arrive AI to enable last-mile delivery solutions domestically.

On data privacy, Ottonomy does not store sensor or environmental data from customer locations. Instead, it relies on behavioural learning derived from robot performance metrics, keeping operations compliant with data protection regulations across its markets.

The company currently runs a fleet of 50 robots and claims to have orders for 500 more. It plans to deploy 200 robots in 2026 and the remaining in 2027. Revenue target for 2026 is $5 Mn.

"Our Ottobots are zeroing in on one critical white space: indoor-outdoor logistics. There is hardly any player doing end-to-end indoor-outdoor. We are picking up blood samples from level two at a healthcare facility, taking the elevator, coming down to the ground floor, going through the access door, travelling half a mile outdoors, and delivering to three labs. That is why we are able to give the highest ROI compared to anyone," Vijay says.