For the first article in a new IEEE Robotics & Automation Society Q&A series on robotics commercialization, we spoke with Markus Waibel, one of the three co-founders of Verity, winners of the prestigious 2026 IERA Award at ICRA 2026.
Founded as a spin-out from ETH Zurich, Verity first became known through Verity Studios, whose indoor drone systems have featured at major live events. Today, Verity applies related technology to warehouse inventory automation, using autonomous indoor drones to give logistics and retail clients accurate, near-real-time inventory data and to inspect warehouse racking.
Waibel spoke about the company’s path from research to live events to logistics, and about the lessons robotics founders can take from that journey.
Can you tell us about your own path into robotics?
I originally studied physics. But during my Master’s, I took a robotics course at EPFL, and that really opened a door for me. Coming from physics, which is a very hard science with a lot of mathematics, robotics felt very different. I remember being shocked by how exploratory it was, and by how much of an art it was, as opposed to a science. That was part of the attraction. Robotics is applied, messy, creative and physical.
After my PhD, I had the opportunity to work with Raffaello D’Andrea at ETH Zurich. I joined with the explicit wish to eventually spin out a company. That company became Verity, which I co-founded with Raffaello D’Andrea and Markus Hehn.
Verity first became visible through Verity Studios. Why start with live events?
The live events business was a deliberate stepping stone. We had worked on indoor drone systems at ETH, and people were fascinated by what they saw. They came to us with all kinds of ideas: cleaning windows, transporting items in factories, logistics applications, and many others.
We were interested in logistics early on, but it was a very difficult market to enter directly. Warehouses are large, complex, cost-sensitive environments. We knew we needed a first market that could help us develop the right capabilities.
A Verity Studios micro-drones display at a 2017 Metallica concert. Credit: Verity
So, we used live events as a stepping stone because they share a lot of important similarities with industrial applications. Safety is paramount, robots need to be autonomous, and the entire system needs to be ultra-reliable. You can’t have a Metallica show with 15,000 people standing there and say, “Oh, we’re almost ready.” There is a moment, and your drones better be in the air.
Live events also made sense commercially. The live events industry is a great space because it pays high premiums for innovation and it creates its own marketing materials because when you do something visually striking on stage, the videos become part of the product story.
Verity Studios micro-drones performing at the Eurovision. Photo: Ralph Larmann
How did that lead to warehouse inventory automation?
The warehouse market was always on our radar because of Raffaello D’Andrea’s previous company, Kiva Systems (now Amazon Robotics). In warehouses, one obvious use case for drones is inventory taking. Warehouses have racks that may be 10 or 15 meters high. Traditionally, a person takes a scanner, gets on a lift, goes up to the relevant location, and scans what is there.
But the sensor doing the scan is tiny. The problem is getting it to the right place, safely and repeatedly. A drone is a very good platform for moving that sensor through the warehouse.
The business problem is also very clear. If a pallet is misplaced, it can be extremely difficult to find. In a large warehouse, the system may say that an item is available, but staff cannot locate it. They say “We know it’s in the warehouse, it’s in our system, we just can’t find it.” That creates cost, customer frustration and operational inefficiency.
What was required technically for the warehouse system to work?
The product is warehouse intelligence, so you need much more than a camera on a drone. The drone has to fly fully autonomously and reliably, all the time, in a busy indoor environment, without GPS. It localizes itself onboard and finds its own way through the aisles, safely and repeatably, in the dark, and between racks that can be very close together.
That flight is the foundation, not the product. The drone moves a sensor to the right place; the value is in the software and the data built on top. Each scan becomes accurate, near-real-time inventory data. The Verity system maintains a digital twin of the facility, reconciles what it sees against the warehouse management system, and flags discrepancies, whether stock is missing, misplaced or recorded in the wrong location. Across a fleet, a central cloud platform keeps learning and improving. That data and intelligence layer is where most of the value sits, and it is what turns a flying robot into a warehouse intelligence system. It is also what the wider field of physical AI is short of. AI is learning to act in the physical world, and the binding constraint is no longer only the algorithms; it is access to high-quality real-world data, captured reliably and at scale. A system like ours produces exactly that.
Once you have that foundation, other use cases open up. One example is rack inspection. Warehouses use heavy equipment such as forklifts around steel racking, and damage can occur. Inspecting that racking is repetitive and demanding. For humans, it is very hard to keep looking for small defects in similar structures, aisle after aisle. People go operationally blind. Repetitive inspection tasks like these are often good applications for robotics. And the underlying capability, an autonomous platform that brings a sensor to exactly the right place and turns what it captures into reliable data, is not specific to warehouses. Any large, complex site that has to be inspected and monitored over and over has the same core problem, from manufacturing plants to data centers.
A fully autonomous Verity Series 3 drone acquiring inventory and facility data in a live warehouse without GPS. A single pass captures in-depth, geo-referenced data for multiple inventory, safety, and inspection applications. Credit: Verity
Moreover, safety and reliability are non-negotiable and have to be designed in from the start. You don’t build a car and then say, ‘Let’s make it safe now,’ and the same is true for a robot working in a real operational environment.
What did Verity learn about product-market fit?
Founders need to develop conviction, but they also need to remain honest about the market. This early phase is really the core thing that the founders need to figure out: product-market fit. And then they need to form a conviction, and then they need to go for that conviction.
The hard part is knowing whether you are being persistent or stubborn. One of the hardest questions is always: should you continue trying and pushing the boulder, or should you find another boulder?
There is no simple formula for that. Persistence matters, but so does choosing the right problem. In robotics, because the technology is so hard and engrossing, it is easy to keep working on the robot while missing the deeper market question.
What should robotics founders understand about fundraising?
The answer depends on the stage of the company, but at the early stage, investors are often investing above all in the team.
That is especially true in robotics because building a complete product is expensive. Most first-time founders cannot build a fully finished robotics product before raising capital.
One common weakness in robotics startups is a lack of commercial experience on the founding team. Many companies in robotics and automation suffer from a familiar problem: “three brilliant PhD students, but none of them have any commercial experience.”
That does not mean every founder has to be a businessperson from day one. But the company needs access to market knowledge early, whether through a founder, a very close client relationship, investors, advisors or industry partners.
What role did the local startup ecosystem play?
Switzerland is a phenomenal place to start a technology company – there’s a reason it’s been ranked the most innovative country for 15 consecutive years. Talent density is extremely high. There’s an abundance of startup competitions, grants, coaching frameworks and university support. Sometimes the practical things mattered most. Being able to rent office and workshop space near the ETH environment was extremely valuable, especially for a company with a hardware component.
But there can be too much of a good thing. Companies that fund themselves with grant money for three or four years are probably too far from their clients. Grants and competitions can help, but they are not a substitute for customer learning.
What was the biggest challenge?
It was different in each market.
For live events, it was market size. Only a small number of productions each year can afford to integrate advanced robotics.
For logistics, it was commercialization, and not just for us but for everyone in the space. If you believed LinkedIn, you’d think warehouses are full of automation. The reality is that the overwhelming majority have none of any kind.
The deeper problem is the gap between corporate innovation teams and operational reality. Many large companies have a team whose job is to explore new technology; that can lead to pilots that generate videos, presentations and LinkedIn posts but no operational value. For a startup that is dangerous. You become the show pony for a large corporate, and a show pony is not a business case. What matters is being connected to the people who own both the operational problem and the budget. If you are not, you risk spending your time proving something to the wrong audience.
How important is the first customer?
The first customer matters enormously, but not every first customer is the right first customer. You need a champion, and that champion has to be close to the operation.
In hindsight, one of the things that helped Verity was working with people who had both operational responsibility and a pioneering mindset. They wanted the solution to work because it solved a real problem for them.
The most important thing for you to be successful with your first clients is the team at that site that will be using your system. If that team takes ownership, the startup has a much better chance.
There is also a mindset shift for founders. You will be doing all the work, and that person, that client, that team will get all the credit for it — and that’s a fantastic outcome for you as a business. If the client starts calling them “our drones,” that is a home run.
What do robotics founders most often misunderstand?
Robotics founders are usually excited about robots. Clients usually are not. Our clients don’t have a problem that they don’t have enough drones. They have a business problem.
That is a hard lesson, because inside the robotics company the technical milestones feel huge. The first time the drone flies reliably, everyone celebrates. But that is not the end of the job. You’re not selling flying drones, you’re selling a solution.
The client does not care whether the problem is solved by a drone, a software patch, a process change or something else. The client cares that the business problem goes away.
What final advice would you give to the next generation of robotics founders?
People assume the hard part of a robotics company is the technology. The technology is hard, but the paths we spent the most time on, and got wrong most often, were the commercial ones: how to reach product-market fit, which clients to say no to, how to go to market, pricing. We changed tack so many times. There is no equation for any of that.
We talked to hundreds of clients, and most of those conversations went nowhere. Plenty of people have a use case; the hard part is working out who is worth talking to. Early on we were naive, assuming we could sell top-down and convince a CEO. Even a win was rarely clean. With one of our largest clients, it was a pilot in one country, then another in the next, then another, more than ten pilots before it reached scale.
In a way, that mirrors my own story. I started in physics, which is about as structured and deterministic as a field gets. Then robotics, which I found was far more of an art than a science. Then the commercial side of a robotics company, which is messier still, and that is where I spent my time at Verity. I was not the one making the drones fly or building the cloud platform; the entire tech stack sat with my co-founder and CTO, Markus Hehn. I spent over a decade working alongside five chief revenue officers, on the side where the certainty runs out.
Even after reaching significant scale, there is always another level. Verity is now active in hundreds of warehouses, but the global opportunity is much larger and extends to it capturing high-quality real-world data in many other types of indoor environments. The work of scaling is still ongoing.
For the next generation of robotics founders, the lesson is not to be discouraged by that difficulty. It is to understand it clearly. A working robot is the beginning. A viable robotics company is built when that robot solves a real problem for real clients, in the real conditions where they operate.
Emmet Cole, Science Communicator
