I used to work on Mars rovers and now I develop self-driving cars. Here's how I made the pivot from NASA's JPL to Amazon's Zoox.

• Olivier Toupet started his career working on autonomous helicopters.
• The former supervisor at NASA's Jet Propulsion Lab now works at AV startup Zoox.

This as-told-to essay is based on a transcribed conversation with Olivier Toupet, 42, a principal software engineer working within the planning and controls team at Zoox, the autonomous-vehicle startup owned by Amazon. Toupet is based in San Diego County, California. He started his career as an aerospace engineer and has an M.S. in aeronautics and astronautics from MIT and a M.S. in aeronautical engineering from ISAE-SUPAERO in France. The interview has been edited for length and clarity.

France is where I grew up and went to school for aerospace engineering.

My original passion is really in aeronautics and airplanes. I got my private pilot license. That school was a top space school in France, and has partnerships with top schools in the US, so I decided to do a double diploma at MIT.

I've always been very interested in the US. I came during high school for a full month to a school in Philadelphia. I was excited to come again, this time for graduate school.

I was lucky enough to find a French professor doing research on aerobatic helicopters.

That was in the early 2000s. Those were small-scale helicopters that were fully autonomous and flying the sky, doing the craziest aerobatics. Very interesting from a trajectory planning and control perspective.

I've always loved mathematics and I've been pretty good at it. That's how I got into controls, motion planning, drones, and autonomous vehicles.

As a foreigner, it's hard to work in the aerospace industry.

There are lots of restrictions. But I was lucky enough that my advisor had a small startup where he was actually commercializing some of his helicopters. The co-CEO was also an MIT research engineer. I got hired by a company, Aurora Flight Sciences, in November 2006.

The first six years of my career, I was working directly with a lot of faculty at MIT and doing really exciting research projects on multi-vehicle autonomy, so drones, ground vehicles, water vehicles, even space with NASA.

As you can imagine, very exciting intellectually, but sometimes it's a lot of small projects where you do small prototype developments, feasibility, and it doesn't really make it to an actual product.

It left me a little bit wanting in terms of having a bigger impact and actually developing real products.

I moved to California right after I got married in 2011 and worked on self-flying cars at a company that's now called Wisk.Aero.

I got to work on some pretty challenging problems. This was actually developing a real airplane from the ground up. So that was a little bit of a dream, a bunch of friends in a startup developing an aircraft.

It's a small world in robotics. I've never actually applied to a company. Of all the jobs that I've had to date, it's been friends reaching out to me to come work with them at that company.

In 2011, it was still a little bit early for self-flying cars. I think now they're starting to gain more momentum, but still probably a decade off.

Another friend of mine reached out and told me I should come join them. I decided to move to LA and go work at NASA. I figured after all the startups, why not work for a big company, and what better big company than NASA?

I'm really interested by space and aerospace engineering, but my real passion was more airplanes.

I was the supervisor of the Robotic Aerial Mobility group at the NASA Jet Propulsion Laboratory.

It is kind of a unique place in the sense that it's called a FFRDC: Federally Funded Research and Development Center. It's at the intersection of industry and academia in the sense that you have some really cool research projects, but you also have really cool NASA flight missions, where you can actually infuse some of those robotics technologies.

For example, I developed the self-driving for the Perseverance rover, so now it can drive itself, on Mars, beyond the horizon. I got a medal from NASA for my work on traction control for the Curiosity rover, which helped basically reduce damage on the wheels. (Toupet received another NASA medal for software for the Perseverance rover.)

I got to drive the Opportunity, Curiosity, and Perseverance rovers, as well as fly the Ingenuity helicopter on Mars.

I had a good friend who left JPL to go work at Zoox.

I joined Zoox as a consultant in March 2015, and that was kind of ideal for me because I got to do cool stuff for NASA at a fairly slow pace — those space programs take many years to get to maturity — but at the same time, most weekends, I would be here in the Bay Area working at a much faster pace. It was kind of best of both worlds, and so that went on for seven years.

Zoox was acquired by Amazon. It was perfect timing.

One reason why it worked so well for me to be a consultant here: When you think about it, somebody may see it as just a few hours on weekends and so on. How can you be productive? The reason why it works so well is because there's a lot of technical expertise that I can leverage and that helps me develop things much more quickly than someone who would have to rediscover how to do things from scratch.

I joined full-time in November 2021.

Connections are very important.

It's just all those strange coincidences in my life that ended up panning out. I don't know if I did anything special, except just following what I'm really excited about. To me, what's most important is the people with whom you work. If you look at my career, even though I really like airplanes, then I went to space and I went to AV, and it's not necessarily because I was that excited about AV or space, it is also because of the people I got to work with.

There are other AV companies, but I would never consider working for another one, because to me, the people at Zoox, they are like family.

The leaders of the company, CTO (Jesse Levinson) and CEO (Aicha Evans), are people that I look up to and that I admire. They're not just people who pay me every month.

The AV space is the next frontier of robotics.

I've worked on lots of autonomous vehicles, but always in some fairly small scope. What we're trying to do here in the AV space, it's extremely multidisciplinary. It's the entire robotics stack. You're doing perception, you're doing planning, you're doing prediction, doing simulation. Zoox is also doing the vehicle as well, so all the hardware. It's redesigning the whole robotaxi from scratch.

If we are able to develop that technology that can react faster than humans and doesn't get drunk and doesn't get tired and doesn't get sleepy at the wheel, that's a fantastic outcome of the robotics expertise I've developed.

The gain is proportional to the risk you take. If you always try to do something easy, then your accomplishments are not as great. I do think that technology for self-driving cars has some challenges, but we will get there. It's a frontier that we will reach soon and I think it's absolutely worth the challenge and the hard work.