Sometimes things don’t go well the first time… or the second… or even the hundredth time. But as the saying goes: if at first you don’t succeed, try, try again. In this week’s episode, both of our storytellers share stories about going back to the drawing board and trying again.

Part 1: As the team principal of the autonomous racing team, Madhur Behl is determined to win.

Dr. Madhur Behl is a professor in Computer Science at the University of Virginia, where he conducts research in Robotics and Artificial Intelligence. His primary focus is on advancing the frontier of safe autonomous systems capable of operating seamlessly in the real world. Dr. Behl is the team principal for the Cavalier Autonomous Racing team, which proudly holds the title of the leading American team in the global Indy Autonomous Challenge. Additionally, he is co-founder of the F1tenth program which has hosted numerous 1/10 scale autonomous racing Grand Prix events, engaging over 70 universities and 1000s of students worldwide.

To hear all the stories from our show with UVA check out our friends over at HOOS in STEM.

Part 2: Engineer Mate King searches for a plane that crashed in the remote mountains of Idaho.

As a second generation Mechanical Engineer (in training) Mate King has spent his life asking "How can I make this work?" Whether trying to figure out how to make a motorcycle ride on water, to designing fuel delivery systems for rocket ships, his world of engineering is just getting started. His dream is to work in the outdoor space, designing cutting edge gear that allows the greatest performance possible in the most extreme conditions.

Episode Transcript

Part 1

Seven years ago, I had a crazy idea. Just like how motorsport racing has always been a proving ground for automotive technology, can autonomous racing be a testing ground where we can improve the safety of self‑driving technology?

I wasn't proposing that these autonomous vehicles have to race through traffic, by the way, but, surely, this chaos of racing and the overtaking provides an opportunity where we can enhance the capability of these cars to handle some unexpected traffic situation. Trust me. I grew up driving in India, so I know a thing or two about uncertain traffic situations.

With this mission in mind, I started the Cavalier Autonomous Racing Team in 2017 when I joined UVA. And you can imagine my pride and joy when, a few years later, my crew of just five students and myself, we emerged as one of just nine universities in the world who qualified to get a chance to compete with a real‑world, fully autonomous Indy race car as part of this global competition called the Indy Autonomous Challenge.

Madhur Behl shares his story at Carr’s Hill at UVA in Charlottesville, VA in April 2024. Photo by Matthew Webber.

This is a dream come true. I pack my bags. I tell my students stop everything and we are moving to Indianapolis, the racing capital of the world.

We lived there for three months to work on our car. Every morning at 7:00 in the morning, I drive my students to the racetrack. As you enter the track, there's a board and it says “General Public” with the arrow facing right. And right next to it is a board which says “Racing Teams” with the arrow facing left.

And each day, when I made that left turn, a little smile appeared on my face. I was at the Indianapolis Motor Speedway, home to the Indy 500, one of the most iconic racing spots there is in the world. And this track is going to be my research lab.

Racing without a driver is not easy. Our software has to operate like an elite athlete. It needs the precision and the skill of a world champion race driver and the brains of a chess grandmaster. There is no driver behind the wheel. In fact, there is no wheel or pedal in the car.

I take a look around in pit lane and the competition is fierce. I see formidable teams from Italy who have dozens of students on their crew and their Italian racing heritage and pedigree on the line.

There's a team from Germany, a university which literally has an entire department of automotive engineering. I have a feeling that this is not going to be easy at all. But as it says, we choose to do this not because it is easy but because we thought it would be easy.

And on one cold winter morning in October 2021, our racing driver comes to life. Our race car goes at 125 miles per hour under no supervision. That's impressive. But it was only 4th fastest overall. We couldn't really win the competition, but it was fast enough to earn us the title of the top American team in this competition.

This lights a fire underneath me. I now believe that I'm not here to just participate. I can contend. I can fight for that podium. And so I carry this confidence with me into the next race.

And this is going to be my complicated because, now, there's going to be two cars on the track with the complexity of how do you autonomously overtake an opponent. It's not enough to be fast anymore. The car has to figure out where the opponent is. It has to predict what their next move is, and it has to find the opportunity to choreograph and go for that overtake.

This time, I travel to the entertainment capital of the world and arrive at the Las Vegas Motor Speedway. We seem to have picked up from where we left off in Indy. In a week of practice, my car is already traveling at 100 miles per hour, going on the front stretch, lap after lap.

I find myself again in a familiar position in the top four bracket, me against three European teams. This time I can do it, I feel.

And just one day before, before the final race, I see my race car accelerate up to 120 miles per hour, overtake the Italian team in a practice session, and just before it was about to disappear into turn one, I hear a loud thud.

I look at the horizon. A plume of dust emerges and the entire pit lane is quiet. The silence is only broken by a call on the radio, “Car Nine has crashed into the wall.”

My heart just sinks. This adrenaline that I just experienced of watching my car overtake completely overshadowed by this severe gut punch.

I get into the safety truck. I drive to turn one, and I know there's no human driver who have gotten hurt, but it feels like my autonomous driver is severely injured. The left side of the car is practically destroyed.

And it's not just the carbon fiber which has crumbled, but my hopes and aspirations of being on that podium have crumbled with it. It was a reminder that when you operate at the frontier, sometimes you have to take risks because it's the unknown.

There's a saying in racing that if everything seems under control, you're not going fast enough. We take meaning from those words to motivate us that we learn from this mistake and try to come back.

Let me tell you something about race day. It's especially disheartening when you don't have a car to compete on race day and yet you are forced to stand in your pit box and just smile and wave at the cameras. It's a feeling I especially dislike.

The next year, we have two more races, two opportunities to come back, but in both of those races, we fail to even qualify. Our once pioneering Cavalier Racing Team is struggling to even go at 70 miles per hour. The steering is constantly trembling while all the competition is making remarkable progress.

Two more times I'm just standing and waving. It's almost become this gesture of defeat for myself. On the outside it must have looked like this is a fading team. They had their moment and now they can't come back. But, internally, we are far from defeated because we believe in the science.

We are constantly improving, revisiting our mistakes. We are not just tweaking, we are making overhauling changes to our software. Every single line of code ever written from the beginning is under scrutiny. And a constant thought lingers in my mind. If only somehow I can get more practice time on the racetrack, we might be able to turn the tide.

In 2014, I get one more shot to return to Las Vegas. This time I feel I'm better prepared than ever before, but nothing in racing is easy, because we only have four practice days before the final race. It's the shortest calendar they've ever put out in three years. Our backs are against the wall one more time.

Outwardly, I'm projecting confidence to my students, but inside, I would be just happy o qualify. After an absence of three years, just being on the start line would have been a victory in itself.

Let me walk you through how daunting this qualification is that I keep talking about. You have to demonstrate two things. The car has to accelerate at 120 miles per hour and then slam on the brakes to show that it can come to an emergency safe stop, fully autonomous.

Next, you have to overtake a ghost obstacle, a virtual opponent. I know it sounds like science fiction, but we have to do this with the real car. And we only get a couple of shots every day because track time is precious. We have to share it with every other team.

The first day of practice, we use up a session to just shake down and make sure everything is in order. And that same afternoon, we let the car go to 90 miles per hour. It looks stable. The steering is stable, like a surgeon's hand.

So we go to 100 miles per hour. My students look at me. They don't say anything. I know they're asking for my approval to go faster.

So I say, “Let's go 110, see what happens.”

The car is still steady. My heart at this point is racing probably even faster than the car is going but I find that confidence inside me. I believe in our effort and I make the call. “Let's go to 120 miles per hour and request a red flag to bring us to a stop.”

The car is now coasting at 120 miles per hour. Now, it's serious business. It's doing a lap in almost half a minute.

Three, two, one, Red flag issued to Car Nine. My students and I are zoned in into our computer monitors. We're not even looking at where the car is on track. We see the car stop gracefully on a dime, exactly how we had predicted.

And just like that, on the very first day, we are able to check one of the two qualification requirements. This gives us confidence.

The next day is about facing ghosts, literally. We take a slow and cautious approach, but our car is able to go at 100 miles per hour and flawlessly overtake a virtual opponent, demonstrating its ability that it can plan around an obstacle.

We check off another qualification requirement.

Within a span of two days, I go from unseeded to the top‑seeded team this year in this competition. Surely, that's great. We've gone further in the competition than we've ever done in the past three years. However, this is not even half of what is required to race in two days' time.

Because the real litmus test is, can you overtake a real opponent? Can you overtake a real autonomous car?

The next day, the plan is simple. We go out with a Korean team just to try to see if we can detect them in order to overtake them. I send out the car. It catches up to the Korean team, but it cannot see the opponent. Something has gone drastically wrong.

I don't hesitate. I bring the car back into pit lane. And before it even arrives in the pit box, I see my students huddle together. They are already deliberating, brainstorming, what could have gone wrong? Why did this not work? What do we need to fix?

In that moment, I choose to step away. I trust that they will be able to figure it out under this time pressure. I let them own that moment.

Much, much later my undergraduates would educate me and tell me that when somebody is in the zone and you let them figure it out, the correct Gen Z terminology is that they were cooking and I'll let them cook.

But the race organizers are insisting we have to demonstrate this overtake, so we fix what we could. We don't know if that has fixed the problem. Is it going to work? Is it going to crash into the opponent? One way or the other, we are about to find out.

So with bated breath, I send the car out one more time. Flashbacks of that crash from 2022 are constantly replaying in my mind because it's the day before the race day and the margin of error is very thin. Something would go wrong, I'll get eliminated again.

And the moment of truth arises. When our opponent is going at just 45 miles per hour and our race car creeps up to them, veers out, accelerates to 60 miles per hour and executes its first ever autonomous overtake, my pit box erupts with the loudest cheer I have ever heard in my life.

I find somehow some strength to compose myself and tell my students, “No, let's keep focus. The car is still on the track.” But internally, I'm screaming even louder than them.

I take a look around, all the other teams are sort of confused because this is a very slow unremarkable routine overtake, and they're not sure why the Cavalier team is celebrating, but they don't know what monumental moment this has been for us. This is not just a slow overtake. It's an overture to the perseverance of our team. In that moment, I see my students grow from uncertainty to confidence. They have evolved from learners to become leaders.

At the same time, I don't want to tempt fate a day before the race day so I just bring the car back and we'll see what happens tomorrow.

It's race day now. Race day is always exciting.

Because we qualified first, our very first matchup is against an American team, a team from MIT. This is all‑American showdown is about to go down.

Head‑to‑head racing works in a manner where you take turns overtaking each other, and if both of you succeed then you try again at a higher speed and a higher speed. It's like a game of chicken, who will give out first?

Our racing car is able to outmaneuver the MIT car at every turn. We can overtake them at every speed, reaching a top speed of 143 miles per hour, the fastest we have ever gone in this competition. And just like that, we advance into the finals.

We are not just creating personal records in history, but we've already secured a place on the podium, which seemed impossible.

Next up is the final race against the German team, a former winner of this competition. I know that the German team will go the distance, they are capable, and to be honest, I feel I'm not going to back out. I feel ready. I'm going to send it, like they say in racing language.

So it's our turn to overtake the Germans, but the Germans are sneaky. They don't make it easy for us to overtake on a straight. They take a racing line which forces our hand that the only way we can overtake them is in the middle of a risky turn.

But our driver has also learned how to race. And so, we execute an overtake maneuver in the middle of a turn at 100 miles per hour, overtaking the German team. Now, the ball is in their court. Can they return the favor?

So the German team lines up behind us, next lap. And just when they are about to overtake, they're literally by our side, my AR driver decides to race away from this overtake. It speeds up.

Unfortunately, that's a rule infringement. The defending car has to hold its speed, just for safety reasons.

I look at my student and he immediately looks at me and says, "Oh, Madhur, I know exactly what's going on. The cost of the prediction is not tuned well."

In plain English, what he's saying is our race car thinks that the German team is about to crash into us, so it performs a safe evasive maneuver by speeding out of this overtake, whereas it should have yielded.

I know we will not be able to fix this right there and then on the track. If only I had that extra practice there, I would have been able to fix this problem as well.

I tell race control, they can try again. Probably the same outcome will happen. That's exactly how it plays out.

We put up a show for the cameras. The German team crawls up beside us, but the Cavalier car is not in a mood to yield today. It races out of that overtake again.

Unfortunately, we get DQed, disqualified, and we have to secure second place in the competition.

And despite the finals not going as I had hoped, I'm immensely proud of what the students have demonstrated. They are able to showcase the true capability of the team at such a global prominent stage. Moreover, we were able to show that the adrenaline in the competition, the entertainment of sports mixes very well with the research, the curiosity, and the problem‑solving of STEM.

As for me, I don't think of myself as just being a computer scientist anymore. In fact, I'd like to claim the title of being the coach of the university's fastest racing team. No American team has yet to win a race in the Indy Autonomous Challenge. I'm fortunate to be working with the students that I have. I very much would like to change that.

Thank you.

Part 2

I remember laughing with my two best friends, Matt and Caleb, as we dug our stuck SUV out of the snow. We had taken a fork off the main road, hoping to get a shortcut, and our nicely groomed road had quickly turned into just frozen, crusty mess, and we were buried up to the axles.

Thankfully, we had the right tools that we needed, and after a couple minutes of digging and pushing, we were back on our way. The total time that it added to our trip, maybe half an hour. But it didn't matter. We had a mission. I had a mission.

At the age of 12, my parents were looking for something to keep a lonely kid who had moved away from all of his friends busy and maybe learn some good life skills in the meantime. I was already in Boy Scouts run by our church group and I had always kind of been sort of a social outcast in that sense.

Mate King shares his story at Boise State University in Boise, ID in November 2023. Photo by Joseph Rodman.

And so, in looking for something new, one of my mom's friends suggested an organization called the Civil Air Patrol. I went to a local meeting that we found and I was instantly hooked. We learned about aerospace engineering. We learned about leadership principles. And then a couple of weeks into my time at the squadron, I found what would become an obsession, almost, of mine for the next few years. We were putting together an urban search and rescue team.

A team. That was something that I was so excited to be a part of and the thought that I could be a participant in it and contribute in such a way was huge to me. So after many years of training and time, I was able to be qualified to a point to be put on our standby mountain search and rescue team.

Now CAP, as it's called, has the nation's largest civilian staff search and rescue air operations. They have a private fleet of airplanes that they use to track the downed black boxes from airplanes that crash and then supplementary teams on the ground to direct EMS crews to the crash sites.

So here we were. The day before, December 1st 2013 we'd gotten a call from the FAA. An airplane had crashed in the remote mountains of Idaho in a narrow town called Yellow Pine. I was part of the team called out to go and try to find this airplane.

It was a huge experience for us, so we woke up early and we just made our way out there. We got through the cascade, made a local base of operations with the sheriff's department, picked up some nice sandwiches that they made at the Shell station up there and went on our way to Yellow Pine.

Minus our little detour, when we got into town, the mountains in the twilight just surrounding the airfield looked enormous. It felt overwhelming to know that we were that point of contact.

We met with the EMS crews already on the ground and we were able to go and coordinate our next day from the information that they had. And we also had the opportunity to meet the family members of the people involved in the crash.

We were excited and the feeling that the family had was so impactful as they expressed to us how grateful they were that we were there on such short notice and also that we were really the only hope that they had to be able to find this airplane at the time.

So, after making our plans, we went to bed. Well, kind of. After an adrenaline‑filled night of talking and not a lot of sleep, we got started early the next morning.

We made our way on to the mountain and the places that we'd planned and things were tough. They were not good. The night that we were there, it had snowed, and the night before it, it snowed, and we were looking for a white airplane with a big purple stripe on the side.

To compound with that, the radio signals from the airplane's black box had been bouncing around across all the mountains and valleys around us. So we'd spend time plotting a course, trying to find the strongest signal that we could, and then, as we moved forward, it would unexpectedly die out. It was frustrating.

But after a whole day's work, we were able to cross off large sections of the mountain and come back for another meeting.

The mood was a little more somber that night, as we all talked together and made plans for the next day. We were able to cross out the sections that we had worked through along with the other crews that were up there and we went to bed again, ready for another day.

We woke up early again, ready to go, selecting new areas, and having basically the same experience as the day before. I remember just being so, so frustrated. I sat down for lunch and I'm making myself a chunky peanut butter‑Nutella sandwich. I was just sitting there. I knew that I could do this, that the tools that I had, it worked. I'd proven them time and time again, and I just felt that if I was able to close my eyes, I could spin around, I could find it, except I knew that I'd probably break an ankle because of all the snow.

So we finished the second day with the same lack of success as we'd had. And as we met with the family members and other EMS crews, it was just very grim. Because of the weather, because of the terrain, our hope was going away.

We started the third day earlier than the other two and in our best efforts, because if we hadn't found anything by midday, we would have to leave.

So, me and my partner split up a little bit more than we had the days before. And going to this new area of the mountain that we hadn't been to, we were able to crisscross in a grid pattern and do about six miles in a few hours, doing the best that we could in snow that was anywhere from ankle to hip deep.

As we were going on, coming to the last few hours that we would have, all of a sudden, I saw something that I hadn't seen in those two previous days, something that was so hope against hope, almost.

Looking above me and to the left, I saw a row of trees cut off at perfectly the same height. It's almost as if an airplane had hit them on the way down. Using my radio equipment, I was able to see that there was a strong signal coming from that direction. So as I looked up over the ridge, I just began to run, doing my best to get to the top and hopefully find something.

As I crested the top of the ridge, right under, all these broken‑off trees they stopped, and so did the signal. There was no plane and nothing on the other side.

There's a very dejected 17‑year‑old who walked off the mountain that day, who's really confused. I had the right tools and I knew that I could use them because I've proven that in the past.

And that is so like being in engineering. The college that we're going to, the classes that we're taking, their tools and how we apply them is how we find success. The thing that needs to be remembered is that we don't control the environment that we're using these tools in, and sometimes you may not find the things that you want as you go through it.

Take my college journey as an example. Boise State is the third college that I've transferred to in my time. I started pursuing an engineering degree back in 2018*, right before COVID wrecked everything a semester‑and‑a‑half into my journey. And since then, there have been so many of these false signals that I've tried to follow as I've been trying to find that path.

But unlike my time on the mountain, I don't need to go home after three days. I can come back day after day, week after week, and semester after semester, although I really, really hope it's only four more semesters. And I can learn. That's the beauty of engineering. Every single day that you come, there is an opportunity. An opportunity to create, an opportunity to learn, an opportunity to grow, and I love that.

Thank you.

*Correction: Mate King actually started school in 2019.