Building an Interdisciplinary Career: From Aerospace and Big Companies to Robotics and Start-ups

Interview with Isaac Brown, Starsky Lead Hardware Engineer

It’s been three years since Isaac joined Starsky Robotics as the first employee. Now, he leads a hardware engineering team and works on what he is really passionate about — making capable and reliable autonomous trucks a real product to roll out on the market. We asked Isaac to share his exciting story including his job at NASA, a desire to work in a start-up, Starsky’s first prototypes along with the further development of its hardware platform, and a one-year sailing trip on a traditional wooden schooner.

Isaac, let’s start from the very beginning. How did you become a hardware engineer? Is it something you’ve always been passionate about?

I didn’t really understand what an engineer was when I was in high school. Maybe now, with tech being more of a thing, everyone reads the news about start-ups and understands what an engineer does. But when I was in high school, people would just say “Oh, you like science and math. So, you should go into engineering.”

I wasn’t sure if I wanted to do mechanical or electrical engineering. So, I did both the electrical and mechanical intro courses at school and I went with mechanical. It’s kind of funny that I did both those intros because I feel like I’ve had a career which is interdisciplinary between mechanical and electrical, and now I just call myself a hardware engineer, to try and be all-encompassing of everything that isn’t code.

You started your career in the aerospace industry where you were involved in pretty unique projects. What were the major highlights of that experience?

Yes, I worked at the NASA Jet Propulsion Laboratory. It was the first place I worked full-time, seeing how engineering is done. I did tests related to the Mars Science Laboratory (MSL), later known as Curiosity.

I was part of the testing group, which is about building custom solutions to support testing. You can’t just send up a replacement satellite or Mars rover, so you need to make sure it’s going to work the first time — not just that it won’t break, but that it’s going to do whatever it’s supposed to do in terms of science. So there’s lots of testing to confirm that it’s meeting all the requirements, and we would build custom equipment just to make it possible to do the tests needed.

Some tests that I supported were done in a massive thermal vacuum chamber. It was the size of a building and could hold the whole satellite at once. It had a sun simulator and really simulated what it’s like in space. That’s completely unique.

How did you make your way from the aerospace industry to robotics and AVs?

I majored in mechanical engineering, but I also did a minor in robotics and I worked at a robotics lab. Robotics is essentially things that move, and for this you need to have mechanical, electrical, and software components all working together.

I reached a point in the summer of 2016 when I had worked at a few different places and industries — from aerospace to semiconductor fabrication. But I also had a huge desire to experience what it’s like working in a start-up.

I came up to San Francisco, interviewed with a bunch of companies and ended up deciding to work with Starsky. Starsky’s particular approach to autonomy with trucks made a lot of sense to me from the business side and I was excited to continue developing my skills in robotics. At many other companies, it’s really more of a software project with a little bit of electrical, where you’re basically taking drive-by-wire vehicles — meaning they already have all the sensors and controls that you need — and just creating software to control them.

When I joined Starsky, drive-by-wire trucks didn’t even exist. If you wanted to be able to control a truck, part of the project was putting your own motors or actuators on the vehicle to spin the steering wheel and push on the pedals. That’s how it fit in with my robotics background within the autonomous space.

You joined Starsky when it was in its early stages. So you could see its evolution from the idea to the fast-growing start-up with an extensive engineering team, multiple locations, and autonomous trucks driving on the roads on a daily basis. Is it exciting to see all these changes?

Yeah, I was the first Starsky employee when I joined in June of 2016. At that time, we had no trucks, so early on we were doing our tests in minivans. We were creating these really rough ways to strap a motor onto the steering wheel of a minivan, which half the time was a rental.

Then that first summer, we got permission to go to a truck yard in Northern Georgia and test our stuff on a truck each day for two weeks. It was basically the same thing we’d been doing with the minivan, and again nothing could be permanently mounted.

Literally every morning, we would show up with a blank new truck. Within an hour, we’d have to put our motors and the whole system on the truck in order to get a good amount of test time. It was all run off of Arduinos and a laptop in the truck, and it was all teleoperation at that time. The progress that we’ve made since that is phenomenal.

What are the major things on the hardware side that you’re currently working on?

The big thing about Starsky is that we are trying to make a real product. It’s not a research project.

We’re not saying: ‘let’s figure out how theoretically we could come up with a solution that would be autonomous 100 percent of the time in all situations.’ It’s more: ‘let’s find the shortest path to getting a person out of the truck while ensuring safety.’ That’s why we did teleop and focused on running in certain locations.

We’ve had a number of iterations that we’ve gone through, almost like continuous integration of prototypes for building up the hardware platform. It has grown a lot throughout the years: adding all the cameras, additional computer power, better design and power distribution.

Now, the main focus of the hardware platform is to package that up into a product. Not just a few trucks that are built out to prove the concept, but more of a batch build of a productized form of it that we can roll out.

Our design effort now is on the productization but also reliability and no single point of failure. For that, there’s some interesting design going on: how do we prevent single points of failure in an efficient way and how do we make the system capable of doing everything it needs to do and make it able to be produced in volume?

You said that before joining Starsky you’d had a strong desire to work in a start-up, and it’s easy to tell that you’re still enjoying it. What do you like about start-ups?

At a start-up, you get exposed to a lot of things, and you know everyone and what everyone is working on. So, if there’s a way you can contribute, you just go help with that problem. That kind of openness is part of Starsky’s culture.

There’s a lot of freedom of ideas and support across the whole engineering team. Every single person here is working towards that same goal.

I think that’s very different from a larger company with billion-dollar projects where an engineer might spend a year designing this one single component to mount this heater that is going to heat this sensor to keep it at the right temperature. There’s no way to conceptualize what’s going on with the whole product and how critical it is for the company. Does this product line even matter? If this product fails or succeeds, does it have any impact on the stock price of the company? Probably not much. But at a start-up, you can really understand how you’re contributing to the success of the vision.

Based on your experience, what kind of background might help engineers break into the AV industry?

On the hardware side, having an interdisciplinary background or a background in robotics generally is a really good way to get the type of skillset that would be useful for autonomous vehicles. Also, if you have general automotive experience either as a major or a specialty within your school, or if you work at an automotive company, that also makes you of interest to any AV start-up that’s looking for hardware engineers.

There was a period of time when you literally paused your career for a year to go on a sailing trip. Could you talk about this experience and the major takeaways from it?

Yes, you might call it a sabbatical. I sailed on a boat for a year. It was a family trip on a traditional wooden schooner. This boat was rescued from the point of collapse by my wife’s family about 20 years ago. They restored it and then we sailed it all by ourselves from Maine to the Caribbean, starting in Rockland and going down most of the Eastern seaboard from port to port as far as Norfolk, Virginia. We sailed to Puerto Rico and went east down the chain of the Lesser Antilles, visiting almost all the islands between Puerto Rico and Grenada. Then we finally sailed back to Maine, also stopping at Bermuda.

While we were on this trip, we also did a little bit of science with an organization called Adventurers and Scientists for Conservation. We gathered seawater samples for microplastic research to measure the concentration of microplastics in different parts of the ocean.

That was part of my life that I was living in the most unconventional way. It probably was the craziest thing I’d done but it was definitely worthwhile. Maybe it made it seem not that crazy to move across the country to join a tiny startup.

***

If you want to join the Starsky team and help us get unmanned trucks on the road, please apply here.