Episode 152 - One Step Closer to L4 AV Trucking

The most polite driver on the road might soon be an L4 AV truck.

Torc Robotics is the first company developing the technology to commercialize fully autonomous semi-trucks in the U.S. for long-haul applications. By running extensive simulations with the best and most experienced truck drivers in the world, Torc’s self-driving vehicle software and integration solution is expected to achieve L4 automation by the end of this decade.

As an independent subsidiary of Daimler Truck, the Torc system uses an approach called See-Think-Act to calculate the optimal driving decision. By employing veteran truck drivers to reinforce their technology with real world know-how, Torc is truly making the road a safer place.

We sat down with Tim Zuercher, VP Engineering – Autonomy at Torc, to discuss the company’s mission to reduce highway deaths, enable critical supplies, and help the transportation industry increase fuel economy, uptime, and capacity.

Meet Our Guest

TIM ZUERCHER
VP Engineering – Autonomy, Torc Robotics

Tim Zuercher is the VP of Engineering for Autonomy at Torc, where he leads the development of perception, localization, behaviors, planning, and control algorithms for Torc’s Virtual Driver. At Torc, Tim has served in multiple roles from Software Engineer to Technical Product Director of the Virtual Driver. Prior to Torc, he was at Embry-Riddle Aeronautical University where he helped grow the robotics research program while developing autonomy algorithms and platforms for land, sea, and air. Tim holds a B.S. in Aerospace Engineering and an M.S. in Mechanical Engineering.

Transcript:

Grayson Brulte:

Hello, I'm your host, Grayson Brulte. Welcome to another episode of SAE Tomorrow Today, a show about emerging technology and trends in mobility with leaders and innovators who make it all happen. On today's episode, we're absolutely honored to be joined by Tim Zuercher, Vice President of Engineering Autonomy at Torc Robotics.
On today's episode, he'll discuss his team's mission to be the first AV company to commercialize a profitable autonomous solution for L4 trucking. We hope you enjoyed this episode. Tim, welcome to the podcast. 

Tim Zuercher:

Thanks, Grayson. It's great to be on.

Grayson Brulte:

I'm excited to have you here because Torc is developing the calm, polite truck. I repeat, Torc is developing the calm, polite truck when you were so kind to put me in the truck. The way it operated like this doesn't drive like a truck. This drives like a very nice, polite driver. How did you make the truck drive so polite and so nice? 

Tim Zuercher:

I think that's a great question. First of all I'm really glad that you're calling out the calm, polite truck. It's actually one of our missions is to exemplify what a driver should be on the road. And to answer your question, we use some of the best drivers in the truck, drivers in the world as role models, and we get them out on the road and we say, Hey, how would you handle the situation?

How would you behave in this situation? We try and come up with the best way to handle that situation and then get our truck to do it. 

Grayson Brulte: 

Look at the million milers. The 2 million milers. Those individuals, the professional drivers. Do you sit down perhaps in a boardroom, we talk about what are some of the craziest scenarios that you've seen out there on the million plus miles, and do you put that into the algorithm? So as your truck approaches a potential edge case that these professional drivers saw, how to react to it?

Tim Zuercher:

Yeah, so on that vein, we've got drivers that have. A phenomenal amount of experience, as you said, a million miles, 2 million, 5 million miles of driving. And we sit down with them and we interview them and we have a back and forth conversation of what have you seen?
How do you handle a situation? How would you handle a situation? We encountered a situation on the road. How should that be handled in the future? Did we do it well? Can you rate us on how we did that? And we get that back and forth going between the, between both our development teams and our operations teams.
To get that feedback in. They've got invaluable experience and we wanna mine that. So it's not quite a boardroom setting, but we do sit down and our developers sit down with them and say, how do we handle these situations? And what's the best reaction from your perspective? 

Grayson Brulte:

Wow. So there's professional drivers as part of the Torc team today?

Tim Zuercher:

Yes. We've got a bunch of professional drivers at our different testing facilities. So when you went for a ride in the truck, you were down in our Albuquerque testing facility and we've got drivers down there that are always in our trucks. So every single one of our trucks that's on the road today has an in-vehicle fallback driver.
And they're the drivers that I'm talking about, those best drivers in the world that are very highly trained to be that safety driver for our vehicle, but also give us feedback. What did we do wrong? What should we do better?


Grayson Brulte:

I want, I wanna highlight your safety drivers. If you wanna use the term safety operators, the level of professionalism that I saw and how engaged they were on the road. You had the one individual in the right seat that was documenting. stuff on the computer, and then you had the professional driver in the driver's seat, and that individual was so glued to the road paying attention to every single scenario. Okay. And he is, you could see, I could see his mind working well, is this car gonna overtake us?

Do I have to take control of the autonomy for a safety perspective? That was really impressive. As you're developing the Torc driver, does that level of safety that you're learning from your professional drivers go into the quote unquote virtual driver? 

Tim Zuercher:

Yes. That's our goal. Our goal is to make the safest driver out there on the road. And our, as you said, highlighting our safety drivers. Our safety drivers are mission critical to us. And they give us that really valuable feedback and they keep us safe. They make sure that we're um, operating the way we should be operating. And you called out that duo. Of the safety driver, safety operator, and then what we have we call a safety conductor.
And they're responsible for the safety of the whole mission, right? So they do the communications, they take the notes they call things out and they look their second set of eyes for our safety operators to really be that pair. We're obviously driving a class eight semi-truck down, down the road, so we wanna make sure that we're on top of everything and we can respond to it as fast as.

Grayson Brulte:

From a safety perspective, I wanna highlight another aspect of the truck. I don't remember the light codes, but when the autonomy was engaged, the lights went on. Yeah. And so when I was in the backseat in a seatbelt, I knew, okay, autonomy was engaged. Those lights went on. Just seems that all these little things went into the truck to ensure the safest operation possible.

Tim Zuercher:

Yes. And that's goes to Torc's safety first mentality. We have an obligation as we're developing this technology to keep ourselves operating in a safe way. 

Grayson Brulte:

When you're having these really safe operations you're testing in a safe manner with all the different things I pointed out with the individual in the right seat taking notes. What type of notes is that individual taking? Is it a pothole? Is it lane markings? Is it perhaps the behavior of the truck? What are those, what notes is that individual taking? And then you as developing the autonomy, how do you utilize those notes to improve the situation? 

Tim Zuercher:

So it really depends on the situation and the thing that the, that we're testing, so there's different types of notes that they take, but they're doing what we call manual annotation. So they're going in and they're saying as you said, Hey, there's really big pothole there. Let me make sure it's called out as a pothole.
But they might also be saying, Hey, there's a pedestrian out on the shoulder half a kilometer up ahead that we've seen. I wanna make sure I get it in these notes to so that we have that annotation and we make sure that we're able to see the pedestrian out at range and that we're responding appropriately to the pedestrian.
So it really depends on what they see on the roads to exactly what notes they're taking, but they're trying as much as possible to give us useful feedback in development so that we can make the best driver.

Grayson Brulte:

You're testing an Albuquerque, which to say the. Is not a very friendly place to drive. It's a complex driving environment. There's beds on roads, there's individuals with, I was gonna say odd behaviors that do things on the highway. Do you feel that testing in Albuquerque is enabling you to build a better Torc driver? 

Tim Zuercher:

Definitely testing in Albuquerque is an adventure. Albuquerque is a fantastic place and from a testing perspective it's even more fantastic.
You have the conjoining of two or the intersection of two highways, so you got I-25 going through, you've got I-40 going through, so it's a freight corridor. You have a lot of interchanges, but then you have a lot of diversity in the types of traffic because you're, you are getting people from out of town.
You're getting a bunch of different types of drivers who interact differently. And if anybody thinks back to the different metropolitan cities, they get in, everybody's nobody drives like they do in Boston, or you gotta drive like a Chicagoan, right? Albuquerque gets a mix of everybody.
And so we see so many different things, but even on top of that, we have a whole bunch of different environments. So I think most people think of New Mexico and they think, desert, it's down on the southwest, right? It's dry, but Albuquerque's actually at a higher elevation than Denver. So we're at really high elevations going into Albuquerque.
We're going through a canyon and there are S-turns going through the canyons. So you're at highway speeds taking S-turns, this is a great environment for testing from that perspective. You have weather, so you have rain, it snows in Albuquerque, we've actually had to shut down operation couple of times just because of the amount of snow that that was snowed.
You have wind, you have dust storms, and then on top of all that you get tumbleweeds. So the amount of environment exposure we get in Albuquerque is just tremendous. So from a testing perspective, it's an amazing place to test. 

Grayson Brulte:

It had that famous wind. That's why the balloons test there, when you're in the Albuquerque Bowl. You have that aspect. It's, it almost could prepare you to drive in Barstow, California when the Santa Ana winds go. It's the, you're getting that environment. And the other thing we'll go, I want to go into the geography, is if you go all the way up to Taos, I don't, your elevation change is pretty steep.

Are you testing the elevation changes of how the truck's reacting as you go up towards Taos and coming down towards Taos, I'm sorry, coming down towards Albuquerque. 

Tim Zuercher:

We're testing all of that. So yeah, elevation changes is an aspect of what we're testing. You're also testing things like, I think most people that have some familiarity with large semi trucks know that they've, they don't just have air brakes or pad brakes, they've got service or engine brakes, right? So we actually can use the engine to, to slow us down. And that's actually a really critical part of driving a truck. You don't wanna wear the brakes out, you don't wanna heat the brakes up, that's not safe. And so those kind of elevation changes allow us to do things like, are we using our engine brake appropriately in our surface brake appropriately when we're testing. A very important thing to do when you're working with a class A truck. 

Grayson Brulte:

What was it like when you first started testing for elevation changes? Because Torc's one of the only long haul autonomous trucking companies I know that's actually testing elevation. Most of your competitors are focused on, let's call 'em flat roads that go for hundreds of miles and they're not really focused on big elevation change.

Tim Zuercher:

I think that's a really great question. I think here the testing here, we've, so Torc is based in Virginia and so we've actually been dealing with elevation for our, pretty much our whole history. And so it's something we're very familiar with and we knew right at the beginning was a challenge that needed to be solved.

Grayson Brulte:

As you look to solve the challenge of, you described the weather, the elevation, Virginia has a lot of weather as well. Did that impact where you put the lidars to the cameras and the radars as you're, let's say, generation truck one to generation truck two to gen three? Do you really look at where's the most optimal place to place the sensors?

Tim Zuercher:

Yes. And I think that's one of the key design elements when you're designing an autonomous vehicle, you want your sensor fields to give you that 360 view. You as a kind of human driver you're, you normally looking forward, you need to make a move to the side. You look to the side and you use your eyes primarily as the driving task.

And anytime you need to look somewhere else, you just go. You look in that direction for the autonomous vehicle, we want to be looking in all directions possible all the time so we can be predicting what's happening. We can determine that it's safe. So the truck says, Hey, I need to make a lane change. Now. It doesn't start looking about the safety of making a lane change. Now it knows that it is or is not currently safe to make a lane change and what the conditions around it are. So sensor placement is a critical aspect of that. And so we wanna optimize not just where we place any one sensor, but where we place them in relation to each other, in relation to the different modalities.

The Torc virtual driver uses Lidar radar. And cameras as sensing modalities. And so we want overlapping fields of view with those different modalities, and then we want to cover in all the directions possible. 

Grayson Brulte:

I noticed when I was in the truck, the prediction were going on the highway and also on this car's trying to come in and the truck knew before the human brain that this person's gonna try and be not a nice person and zoom in front of us and going back to the very polite, calm truck, it slowed down gently, but I didn't feel it as a passenger. How was that possible from a technical aspect? Were you running a prediction scenario to determine, okay, there's a good chance this person's not gonna be a nice driver and we're gonna have to slow down a little bit. How did you do that? 

Tim Zuercher:

Yeah, so there's several different things that can be used to that. So one is the one you ran on is running a predictions scenario. Basically, given the context of the environment that we're in and what the truck is doing and what the traffic around us is doing, what it's, what's the likelihood that this person in traffic is going to cut in front of us. And there's different kinds of scenarios there. Like you can tell if somebody's driving aggressively, and you, you want to, get away from them or let them over. But you can also tell things like in Albuquerque where you took a test ride. The when you're going through downtown Albuquerque, just like any major metropolitan area, there's constantly on-ramps and off-ramps and on-ramps and off-ramps.

So I can watch a car come down an on-ramp, right? And be in that on-ramp lane and know that they're gonna be right next to me and that they're getting onto the highway. And so from that contextual information, the truck can know, Hey, this person's getting on the highway. And it doesn't hurt me to either maintain my speed or slow down just a little bit to give them some space.

And it probably actually helps the whole situation from a safety perspective to just give them a little bit more space. It's not something that the truck is required to do, but even you or I as a good driver, we do that. We're like, we were looking in. We wanna zipper together so that everybody's able to go forward as opposed to just blocking everybody out.

Grayson Brulte:

Grayson Brulte: No, that's well said. And there's another thing I noticed. I don't remember the highway, but it went off to the right and then you went on an overpass and came down and three lanes merged together. Again. The truck was polite and calm and everybody merged politely, and the truck didn't have a hiccup.

It just, it kept going as the car went in front and the car went back. Is that another example of your prediction software? 

Tim Zuercher: 

Yes. The way the virtual driver works is all these different behaviors are layered on top of each other. So they all play together to determine what the best action to take is.

And so what you were seeing there in that specific example was an example of our merge behavior, picking the appropriate place to merge in with other traffic. And then sliding us into that location smoothly. And so it's not so much a courtesy situation because in this case we're trying to enter the flow of traffic.

But all of those kind of behaviors play together to produce that calm, polite truck. 

Grayson Brulte:

The calm, polite truck from a driving perspective. You call this a see, think, act approach. Could you talk about Torc's see, think, act approach please? 

Tim Zuercher:

Yeah. So actually this is not something that's specific to Torc. It's a kind of a thing in robotics. So when you're talking about a robot and an autonomous thing whether it's a Class eight truck or spot walking around in San Francisco, you're the kind of the way to frame what a robot is doing is in that see, think, act or see plan, act basically, Understand your environment, right? And so you have a set of algorithms and methods, sensors for understanding what the environment is, what's going on in the environment, understanding what your position in the environment is, where you are.

We call it localization. And then the next step is that think or that plan, okay, now that I know what the environment is, I know where I am, what's my plan? What do I need to do next? Come up with that plan. And then you have act, act on that plan. And then this is like an OODA loop as a terminology where you loop back and you just do that over and over iteratively.

So from the virtual driver's perspective seeing is ingesting all of the data from all the different sensors. Using that data to localize ourselves, using that data to find everybody else in the scene. And then planning is that behaviors aspect that I was just talking about. All of these different behaviors saying, what should I be doing in this scenario?

Coming up with one plan or one action that we should do. And then we have what we call motion control, which. Acting on that plan and driving us to that plan. And then we do that iteratively. 

Grayson Brulte:

Does that all start a close course testing as you start to build out the philosophy? Do software updates? Perhaps you change where you're gonna put a certain sensor.

Tim Zuercher: 

Yeah, so this all starts with design, right? So the way we wanna develop. Is we want to understand what problem we're trying to solve. What is the product that we want to create? That calm, polite driver, that long haul hub to hub trucking, right?

We say, what's that problem? And then we start saying, how do we start understanding that? What are the requirements to solving that problem? And then we start doing design and we start doing experiments. And part of those experiments, we run them in simulation. So we do quite a lot of. Work to develop things in simulation and to develop better simulations of what you know, of the real world that we can run our designs against.

And so even sensor placement being so important, we even do simulated sim sensor places, sensor placements, and put sensors in different positions and see what their field of view would be, what they would see. If you're talking about a lighter, how many points returned on objects, right? Camera.

How far can I see with this camera? What's different light. Kinds of things. We can do that in simulation, but then we also do that in the real world. And the first step of doing that in the real world is taking into the close course. So in Virginia we've got working relationship with Virginia Tech Transportation Institute, VTT I, and they have a track and it's right next to our headquarters.

And we get on that track, which has elevation by the way, quite a bit of it. And, but we get on that track and we test different scenarios and we test different setups to figure out what. And then we go from there and iterate on it. And sometimes we'll take it to a different track. So we've got other tracks that we use. Around the country to do that kind of closed course testing. 

Grayson Brulte:

You're taking a very, I'll say a smart, holistic approach to testing on the backside of that, are you building in redundancy into the system and the hardware that you're building. So perhaps, I'll give an example.

This has happened to a lot of people's, happened to me. A rock pops up and cracks your windshield on a highway. It's a common occurrence. But in your case, the rock pop up, unfortunately. Crack a lidar, the accounting department's not happy, but the truck still has to get there safely. So is there redundancy if a situation like that happens on the road?

Tim Zuercher:

Yeah, I think that's a really fantastic question. So it's one of the really hard things about level four autonomous driving is that redundancy. So if you think about developing an autonomous vehicle there's step one, which is develop a vehicle that just can drive itself. Steps beyond one, step two and beyond are, it needs to drive itself when things start failing, right?

We have what we call ASIL goals. So there are different ASIL ratings and there's just a standard for how often or how. What the failure rate of something can be and the system as a whole. And our goal for the autonomy system as a whole is the highest ASIL rating, which is ASIL D, which means the chances of it failing are very low.

And so what we want to do to accomplish that is redundancy. And so that's redundancy. Not just in software, but also in hardware. And so that goes back to what we were talking about. Okay, we have cameras, we have lidars, we have radars, right? How do they need to overlap so that I can accomplish my safety goal, right?

Not just so that I can make sure I see something, but I also have to guarantee I can see something if one of those things fails, right? Or if one of those things are blinded, right? So that's from the sensor perspective, but you also have redundancy from a compute perspective. So I. I'm sure, I don't know if it's happened to you, but it's happened to me.

You're driving down the road and all sudden your car just lights up with all kinds of warning lights and it's Hey, please pull over. I don't wanna work anymore. That's usually a compute failure, right? And you plug in your little reader and it says, Hey, replace this O2 sensor are usually, I don't know why , but, that can happen too, right? Anything can fail at any time. And so what we want to do is we wanna make sure we have those redundancies in place, so we want redundant compute, and then you have to distribute our algorithms to those different computers to make sure that if one goes down the whole, the system as a whole can still accomplish its safety goal, which is to be a calm, polite driver and reach a safe position.

Grayson Brulte:

Torc. You have the relationship with Daimler Truck I think it's an absolutely fantastic relationship. The Freightliner Cascadias are world renowned trucks with the relationship with Daimler Truck? Do you get access to the redundant systems in the Freightliner Cascadia, or perhaps Daimler's making a bespoke custom Freightliner Cascadia for Torc with all the redundancy built into it. 

Tim Zuercher: 

Yeah we do have that really close relationship with Daimler Truck and yes we do work with Daimler Truck on developing the system as a whole. So Torc is developing the software aspect of it and to really be safe on road and to really get there, you need a redundant chassis.

We, we need a lot of those hardware parts that you talked about. And so Daimler Truck is working on building that redundant chassis that we can then utilize. For the whole system and uses the solution. 

Grayson Brulte:

As Torc focuses on software, Daimler focuses on making redundant chassis and some of the world's greatest Class eight trucks. Do you collaborate together from a software perspective? If, Hey, Mr. Engineer, Mrs. Engineer, if we added this here, we could have a better performance. Is it a collaborative experience? Collaborative environment. 

Tim Zuercher:

Yeah I would say it is a collaborative environment. We have different levels and degrees of collaboration.

But the closest level we have what we call our Tormlers, right? These are individuals within Daimler Truck who work directly with Torc. And nobody would let me call them Dorkers. So we call them Tormlers. That's good. But that's the, that closest level of collaboration, but that partnership with Daimler Trucks gives us the ability to collaborate at different levels and to discuss what an autonomous truck really needs to be successful. 

Grayson Brulte:

You mentioned earlier when you eventually commercialized, it's gonna be a hub to hub model, I believe it's a white paper on your site, or I've seen public statements on the hub to hub.

How is that approach impacted from the way that you're developing the autonomy stack? Are you developing it differently than somewhere? Or we're gonna go all the way from the warehouse five miles on, we're gonna go on the highway and you say, no, we're gonna go hub to hub. What is the impact from an autonomy development standpoint for that model?

Tim Zuercher:

The impact here is there are different ways you can solve these problems, right? Autonomy and the vehicle being completely alone is just one of those ways. And so you'll see in the autonomy industry there, there's different levels of how mu, how much of the problem are we going to solve, right?

When we talk about a hub to hub model, what you're doing with that is, one, you're concent. Freight traffic. Which is a value proposition, a business proposition for both the customer and for us, right? To have that defined lane right, that you operate in. The other aspect of it is there are different ways that you can get to the highway, right?

And you can control those ways to the highway. Whereas if you're trying to drive everywhere, you don't have control over those and you open yourself up to a lot more things that are a lot more difficult to control. 

Grayson Brulte:

The other aspect is consumers, parents, moms, dads don't want 18 wheelers going into a local community.

When I was in Albuquerque, you demonstrated to me, I think it was, I'll say three blocks, four blocks off the highway. It's a, it was a demonstration that we view this as being a hub to hub and they showed me how your vehicle could do. Surface streets. Is that something? So that was the, I was three to four blocks, I don't remember the exact race, but very close to the highway.

As you're developing the autonomy stacker, you're gonna set a quote unquote range. Okay? We're gonna operate within half a mile of the highway. Are you looking at it from that way? Cuz it's constrained. You're gonna have commercial traffic, you're going to have lights.

Tim Zuercher:

Yeah, we're constantly looking at all of that and we're working with our partners on where does it make sense to have hubs and what should a hub look like and what should be, what should we enable between a hub and a highway? If you talk about deploying autonomous trucks across the entire country, Places in the country look different, right?

And so there's different solutions for each of those places. So I think the quick answer to your question is it's a very iterative approach of understanding where do you want, where do we wanna operate? And then how do we do that as safely as possible? 

Grayson Brulte:

That's a very good point. I've been studying a lot depots or hubs term you wanna use. I've been kind geeking out running all these different scenarios, speaking to everybody to gather as much information as possible on depots/hubs. and I keep coming back to one thing where there seems to be interest. No, not really interest.

I'd love to get your opinion. Will there be technicians at those hubs? So as I described, a lidar gets damaged. Perhaps there's a, you wanna call a Torc certified technician that can replace the lidar, replace the camera. Do you see technicians being there to quote unquote launch the truck perhaps?

Tim Zuercher:

So I think the official ancillary I give you is we don't know yet, right? So this is part of that iterative analysis in understanding what our product needs. The personal anci I give you is probably, and I think personally that makes a lot of sense to, to have a technician there checking out the truck, making sure everything's in order.

We, we working, interfacing with our mission control to say, yep, the truck's good to go. Giving mission control that allowing mission control to give that clearance to yep, take off and go deliver the freight. I think the, my personal answer is probably, we'll see how it develops over the next couple of years.

Grayson Brulte:

How about data? Your trucks generate a massive amount of data. You can't plug an ethernet cord in, can't plug a USB sticker.

They're probably violating the, your security policies. But , you can't do it cuz the amount of data that you get there. Airlines for example, I consulted for a long time in the airline business and they called it sneaker net When they would put the I ffe the inflight entertainment and individuals run out to the plane and plug it in.

Are you developing a sneakernet since? Are you just basically pulling out hard drives, putting it into a wreck? How are you getting all that data off of those trucks? 

Tim Zuercher:

No, that makes sense. And you're right, our trucks collect a tremendous amount of data. And we use that data in development for exactly how we get the data off the truck that's proprietary.

But we do, we get all of that data off of our trucks and then we beam it up to our cloud provider. Then we have access to that data for development and iterating to develop simulations to. To what we call recompute and run different versions of software against what happened. So we make quite a lot of use of that data.

Grayson Brulte:

So perhaps you have sneakernet won't push you on it, don't know, but perhaps when your trucks goes to the depot. The data come off, those trucks get uploaded there in real time, so back at mission control, they can see everything. Or when it comes back, I'll use the term home. When the truck comes home from its mission, then will the data be unloaded?

Tim Zuercher:

That's a, that's a really good question, and I think that's, again, part of an iterative development as you're deploying these trucks. The real question I think here is how much data do you need in deployment? I think it's not sustainable to record everything. all the time from when you think about going from, one truck to a thousand trucks, right?

And you're trying to record and upload a thousand times the data, but then you just also don't need that data, right?

Interesting. Things don't happen all the time. Most of the time it's just normal driving. You're driving down the highway of if you think about driving through Texas, right?

You, if you drive through 800 miles of Texas, most of it's just road, right? And something might happen along the way and you wanna record that and save that. So that's aspect one. What do you actually wanna record? And then as for getting the data off, that's just working with your partners and saying, Hey what's, what makes sense?
For sending that data and there's many different ways you can do that. You can use your sneakernet, you can use over the air to transmit data. There's a lot of different ways to approach that and at Torc, we pride ourselves on, we don't try and say, this is the solution we try and work for. What is the solution that makes sense? And that solution might vary. 

Grayson Brulte:

That ties into what I experienced in Albuquerque is your culture, and I've said this to a lot of individuals, the Torc culture was really, Impressive. It allows you to collaborate as you collaborate throughout the culture.

You're not gathering all the data is one of the premises. Okay? How do we optimize the data? How do we optimize the operations to give you and the team the tools that you need to build the best calm, polite driver in the world?

Tim Zuercher:

Yes. And that's, that's part of what we collaborate on. Are we focusing on the right things? Have we answered the right questions? Are we collecting the right data? And we're constantly evaluating that and saying, do we need to pivot to make sure that we're accomplishing our goals and working incrementally towards delivering that product? 

Grayson Brulte:

You're well down the road to delivering a really great product. What is the future of Torc? You're on the road to success, but what's your future? 

Tim Zuercher:

We are well on the way to on the way to success. I agree with you there. I think Torc's future is to be one of the, if not the main provider of L4 semi truck software to this country, and to, to our customers.

And I see that we're taking an incremental approach to get there. We're working with customers, we're trying to understand what it is they actually need us to solve. Instead of creating a product that then we say, Hey, but you need this. We want to create a product that says, you told us this is what you needed.

Does this fit that? And that's that's our future. And that's our approach. And that's how we plan to go forward.

Grayson Brulte:

You're going forward with an incredible gentleman named Peter Vaughan Schmidt as your CEO. The gentleman has the vision for where Torc's going and to our listeners, I would say what Peter's building and the team at Torc and Tim are building is very special. So be on the lookout for great things from Torc in the future. And Tim, as we look to wrap up this insightful conversation, what would you like our listeners to take away with them today?

Tim Zuercher:

I think I'd like l isteners to take away that one, Torc is hiring. We are actually out there, we're developing really cool technology.
We're developing the next innovation and then probably one of the greatest innovations of our age and come join our team.

We're definitely hiring. And then I want everybody to know that the Torc is a collaborative environment and Torc is looking to develop the best product we can in the best way we can. 

Grayson Brulte:

If you're in the market looking for a job, I can vouch for the culture of Torc. I love what Tim says. They're developing really cool things. After all, they're making trucks drive themselves and autonomous trucks are gonna have a very positive impact on the economy. Go team Torc. Develop the driver.

Today is tomorrow, tomorrow is today. The future is Torc. Tim, thank you so much for coming on SAE Tomorrow Today. 

Tim Zuercher:

Thanks for having me, Grayson.

Grayson Brulte:

Thank you for listening to SAE Tomorrow Today. If you've enjoyed this episode and would like to hear more, please kindly rate review and let us know what topics you'd like for us to explore next.

Be sure to join us next week as we speak with Edward Walker at Hub International. He'll share how the company's facilitating a new era of transportation by providing insurance for the mobility industry. 

SAE International makes no representations as to the accuracy of the information presented in this podcast.

The information and opinions are for general information. SAE International does not endorse, approve, recommend, or certify any information, product, process, service, or organization presented or mentioned in this podcast.

 

Listen to the full back catalog on your favorite podcast platform.

   

X