Mastering robotics, IoT and AI at once - Podcast with Ahti Heinla
2020-10-08 - Ahti Heinlah went through his first computer programming Bootcamp when he was ten years old. This was back in 1982. His mother came home one day and decided to teach him how to program computers. In just a few days, he created his very first computer program. They didn’t have a computer at home, so they went to his mother’s workplace to run the program. It worked. The whole experience excited him and introduced him to the world of electronics engineering and computer programming. In this podcast, Tobias engages Ahti on how he has managed to be successful in three different and complex areas at the same time: IoT, robotics, and AI. Ahti is the co-founder and current CTO of Starship Technologies, a company that’s doing groundbreaking work in robotics. Tobias talks to him about his formative years and his journey to becoming one of Estonia’s most respected founders.
Ahti is best known as one of the founding engineers of Skype, but he’s been part of many startups. He has always loved the idea of trying out different ideas. It is this same approach that pushed him and fellow Skype cofounder Janus Friis to start Starship.
“We were sitting together and thinking, ‘we want to work together: what are the big startups that we could do? What are the ideas that we could work on together?’, he explains, “And we thought, ‘OK, what can we do with robotics that is potentially bigger than Skype?’.
They wanted to venture into robotics because the field was starting to open up for startups and sought ways they could use robotics to touch many people’s lives. Six years later, Starship has grown from “two guys, a napkin, and a whiteboard to a couple of hundred people.”
Ahti was Starship’s CEO in its first four years but relinquished the position to Lex Bayer, instead choosing to be the company’s CTO.
“I actually never wanted to be a CEO. I always wanted to be a CTO,” he says, “Running the whole business: I didn't really think that's my cup of tea, but I can run the technology part of it.”
Tobias engages Ahti on the inner workings of Starship robots, including whether they’re entirely automated. The robots are programmed with several different messages it can play based on the issues that arise when interacting with humans. Still, the automated system can alert human operators to intervene.
Starship is in the process of mapping and annotating the world for its robots, but in places where this has already happened, they have transformed the possibilities in package deliveries. Being the first company to create and roll out autonomous driving robots in public places, Starship has had to build all hardware and software from the ground up.
“We’re building practically everything on our own, one of the reasons being that we started this in 2014. So, there wasn’t a lot of (licensed software) available,” Ahti intimates.
“With robotics, it's also that we realized immediately that we need to do a lot of vertical integration. We decided that we were going to build our own hardware. We were going to design our own hardware. We were going to select ourselves what sensors we were going to use and build the software to work with these sensors.”
Tobias asks Ahti about Starship’s hiring process, given the challenge many companies find themselves in trying to get data scientists and engineers to collaborate.
“It is indeed a recurring theme in our interviews. Generally, the most valuable employees, the biggest rock star engineers in Starship are the people who can do data science, can do machine learning, can write production software, and can build data pipelines.”
What’s Starship’s secret sauce? Ahti says it is its overall complexity.
“It is a very complicated technology that we are building. It's a very complicated business that we are building, and that is actually the sensible competitive advantage against the competitors,” he says.
“We are still the world leader in this field after six years. We've always been the world leader in the field of autonomous delivery robots.”