Employee Spotlight: Lucy Hoag

From building satellites to leading autonomous driving programs

Lucy came from the world of aerospace engineering and holds a PhD in Astronautical Engineering from USC. She’s worked on advanced space system programs at DARPA (ranging from disposable satellites in low earth orbit to a robotic spacecraft servicer in geostationary orbit), and on aerospace engineering projects at Google. Now, she’s a Technical Program Manager for Level 5’s software team where she leads autonomous test programs, helps maintain our AV fleet, and drives improvements in developer and operational velocity.

Lucy is one of many high-performing team members on the Level 5 team who transferred skills from a different industry to become integral in reaching milestones. We asked her to share more about her experience as a Technical Program Manager (TPM) at Level 5, and the steps she took to get here.

How long have you been with Level 5 and what’s your day-to-day like?

I’ve been here a year and a half. As a TPM, I focus primarily on the autonomy software team, which is building the software that drives the car.

No two days are the same as a TPM. At a high level, I work with the product and engineering leads to plan and execute our high level test campaigns and milestones. As we’ve progressed and hit milestones rapidly in the last year and a half, our mindset is now shifting to questions like, “How do we do this at scale?” The team is growing, so we need to be quick, efficient, and intelligent in our execution.

How have you been able to address the challenge of hitting milestones in the midst of intense growth?

One way we describe what we’re doing is that we’re building the machine that will build a self-driving car. This perspective is useful because it helps us remember to invest in sustainable infrastructure and processes to make sure that we’re doing this in a way that’s safe and scalable.

Can you tell us more about what you were doing before Level 5?

I studied aerospace engineering, specifically small satellite design — so I focused on how to optimize and automate the way we design satellites. I spent three years at DARPA as a Systems Engineering and Technical Advisor (SETA), where I worked between the program managers who led high-level programs, and all the different companies that provided technology for the program. Two cool projects I worked on were Phoenix (a robotic spacecraft that services other spacecrafts in geosynchronous orbit), and SeeMe (a low-orbit, disposable satellite designed to see people on the ground). It was high risk, high reward, a little bit crazy, and really fun.

I then went to Google to join a cool (but confidential, sorry!) aerospace project. After that, I transitioned to Waymo for about 8 months before joining Level 5.

How was the transition between aerospace and self-driving? Do you have any advice on transferring skills for folks who may be considering a shift in career focus?

A self-driving car is a little like a satellite on the ground. It uses a lot of the same sensing modalities and principles of operation so this transition made sense to me.

Self-driving cars are also the ultimate systems engineering problem. It’s very unconstrained, and you have to design for a lot of unexpected factors like weather, people, and all the other situations you encounter on the road. So, you have to build this car that’s hyper-reliable, hyper-safe, and has a lot of redundancy (because ultimately, the goal is to build a machine that can sustain itself in the future without needing our intervention).

The core analytical thinking that comes from any engineering background is also applicable for building self-driving cars. No one’s done this before and there is no handbook, so you really just have to be comfortable operating and experimenting in completely new territory.

Any interesting technical challenges you’ve faced during your time here?

I had the interesting challenge of helping drive our program milestones from the very early days until now. When Level 5 started, there were 20 or so autonomy engineers and hardware engineers, looking at each other asking, “What’s the first thing we should do?” It was a really cool to be able to start from scratch.

The first thing we did was sort of a proof of concept — we demonstrated a very simple autonomy stack using a single end-to-end neural net. We trained a model that would allow the car to drive in a circle in the parking lot, only turning right (we affectionately called it Zoolander). The only input was imagery and the only output was steering wheel angle. The system had no notion of lanes, or other cars. This helped us to confirm that all our sensors could talk correctly to each other, and proved that we could execute the end-to-end pipeline from sense to actuate. Then we were able to take that data to troubleshoot, make improvements to the system, and test it again.

What’s the most exciting project you’ve worked on?

When we have a new class of vehicle that we’re building, it’s up to the autonomy team and TPM team to bring each vehicle up for the first time, and incorporate them into the fleet. We’ve developed a rigorous test plan for how we verify the car is approved for autonomy testing. First, we confirm the vehicle’s ability to engage and disengage autonomy reliably, and we test basic low-level controls. Then we verify the ability to localize, and we prove out the perception stack. Then we test partial autonomy, full autonomy on a closed course, and work our way towards public roads.

How do you see self-driving cars impacting people?

There are so many different ways — there’s the safety aspect of eliminating human error, the environmental aspect of having fewer cars on the road, and more. But once this ecosystem is in place, it will be scalable and therefore more affordable. This will unlock and equalize transportation options for people who don’t have the privilege of car ownership or paying for rideshare today.

How would you say you’ve grown since joining Level 5?

Since we grew from a small team, we got to build our culture and define how we produce work, collaborate, and treat each other. That forced me to understand what my values are in this sense and build a team that is effective and inclusive.

What was the outcome of shaping the culture and workflow at Level 5?

One of the most important things, in my opinion, is our focus on keeping very open communication. We have to remain transparent and honest at all times in order to move forward effectively. This can be challenging, but we’re building a very complex system, so it’s critical we can do this.

Any communications tips you might be able to share?

As a TPM, my technique for this is basically just to err on the side of over communicating on all the different mediums you have. By nature of being a very inter-disciplinary engineering team, you’re going to have a diverse set of ways people like to communicate — whether they prefer Slack, email, text, in-person, or something else. I like to communicate “all the things, in all the ways.” At times it can feel like a lot, but I try to remember that if one person asks this question, others may also need the information as well

What gets you excited to come to work every day?

I mean, it’s one of the coolest products you could work on. We’re building this amazing robot that we get to breathe life into. We get to test it to make sure it’s doing the things it’s supposed to do, and when it’s not, you get the chance to troubleshoot it — which is fun. It’s like a different puzzle every day.

Do you have any advice for women in the industry?

It can be intimidating being in this industry and belonging to an “underrepresented in tech” category. My advice is to exude confidence even if you may not feel it. You can’t be deterred.

Another is to remember that you may have different interests or work styles than others on a (predominantly male) engineering team. This doesn’t invalidate your ability to build this highly technical product. It’s important for women to remember — and it’s a key tenant of Lyft — to be yourself and bring your full self to work.

And finally, can you share something about yourself that’s not related to self-driving cars?

I love fostering rescue dogs, and have gotten to work with some awesome rescues in the Bay Area. Something great about Lyft is that I can bring the dogs to work, which is an amazing way to socialize them. The team also loves them, of course!

Lucy’s foster dog with an early test vehicle

Check back soon for more posts in our Employee Spotlight series, and see if there is an opportunity at Level 5 for you in our open roles here.