Autonomous Trucks Are Already Running Freight in Texas

In this episode of Everything is Logistics, Blythe talks with Tete Xiao, VP of Engineering and AI at Bot Auto, about how autonomous trucking is being used in real freight operations.

Bot Auto operates fully driverless trucks in Texas, with current lanes between Houston, Dallas, and San Antonio. Instead of selling software, the company works as a transportation-as-a-service provider, moving trailers for customers using autonomous tractors.

They cover:

• Why Texas and the southern freight corridor make sense for autonomous trucking
• How Bot Auto handles drop-and-hook freight with fully driverless trucks
• Why night driving is one of the clearest use cases for autonomous freight
• How autonomous trucks can operate without driver hours-of-service limits
• Why Bot Auto uses LiDAR, cameras, and other sensors instead of betting on one system
• The difference between replacing drivers and filling lanes that are hard to staff
• How the company thinks about safety, lane expansion, and commercial viability

This conversation is part of the CargoRex AI Use Cases in Logistics guide, featuring real examples of how logistics companies are using AI across freight, warehousing, procurement, visibility, and operations.

Read the full guide here:
https://cargorex.io/research/ai-use-cases-in-logistics/

LINKS:

Bot Auto: https://bot.auto

CargoRex AI Use Cases in Logistics Guide:
https://cargorex.io/research/ai-use-cases-in-logistics/

-----------------------------------------

THANK YOU TO OUR SPONSORS!

SPI Logistics has been a Day 1 supporter of this podcast which is why we're proud to promote them in every episode. During that time, we've gotten to know the team and their agents to confidently say they are the best home for freight agents in North America for 40 years and counting. Listen to past episodes to hear why.

CargoRex is the search engine for the logistics industry—connecting LSPs with the right tools, services, events, and creators to explore, discover, and evolve.

Digital Dispatch maximizes and manages your #1 sales tool with a website that establishes trust and builds rock-solid relationships with your leads and customers. 

Transcript

SPEAKER_00 0:05

There's an estimate of uh 400,000 driver shortage uh in the country. That's gonna happen in the next couple of years. The Department of Transportation, the federal government are uh onto that and the um you know they're aware of of the shortage and that homeless technology is the best solution uh for that.

SPEAKER_02 0:31

Perfect. All right, so let's start at the top with a little bit of your market information. So introduce yourself, name, title, company, and the the target market you're you're trying to go after.

SPEAKER_00 0:44

Great. Uh I'm Tedu Shell. I'm the VP of uh engineering AI at Bot Auto. Bot Auto is a transportation as a service company. And the business model is simple. If someone wants a trailer to be transported from one place to another, that's when you give us a call. The difference between us and the rest of the industry is that we do this uh via autonomous driving. So our trucks are fully humanous. Um, that's how we guarantee the capacity and also able to serve the market that no human drivers are willing to serve.

SPEAKER_02 1:21

And so when you are are, I guess, operating, uh are you operating in a specific area of the country, in a specific area of the world? Uh tell us a little bit about what that that target market looks like that you're operating in currently.

SPEAKER_00 1:34

Oh, we're currently headquartering in Texas, and Texas is our primary market. And the reason is that while uh logistics is a more than actually a $2 trillion market in the United States, you know, shipping via trucks is about like a trillion dollars. Amount of trillion dollars, half of that, so $500 billion, is is done by class eight semi trucks. And 8% of the entire market is located within the state of Texas, and that's where we operate our business. And Texas is mainly like you know, three cities. Uh, we're based in Houston, and we transport between Houston and Dallas and Houston and San Antonio. And this is a perfect use case uh for autonomous driving. And the uh southern corridor uh of the United States, uh, namely, for example, I-10 uh between uh Los Angeles and all the way to Florida is the perfect corridor for autonomous driving, and that's 22% of the entire transportation market.

SPEAKER_02 2:37

And why is it a perfect corridor for that for this target audience?

SPEAKER_00 2:41

Well, because you know, we can go you know nonstop. Uh autonomous trucks can uh, for example, go from San Francisco all the way to Texas within the day because it doesn't have to rest. So you're able to get, like, for example, next day shipping actually, uh, via autonomous trucks. Um, and also, you know, this is a sunshine belt, so the weather is a little bit more favorable. Of course, we're able to operate under heavy wind, have heavy rain, uh, but at least it doesn't snow. Uh, so uh, you know, it's safer to operate right here. And the market is is already huge. Now, once we've mastered this market, of course, we're gonna expand to different corridors as well. But overall, you know, there's a lot of ports uh and and shipping lanes um the southern uh southern uh uh United States, and that's that's a great uh starting point for us.

SPEAKER_02 3:34

Is there a specific commodity that works best for these autonomous routes, or is it really applicable to any commodity?

SPEAKER_00 3:42

Um anything works. Anything works. So, you know, it's an autonomous uh drop and hook kind of model. So you have a trailer, you know, we have a bunch of autonomous tractors, and you know, we go and pick up these trailers, you know, we hook these trailers and off they go uh on their own. And once they uh arrive, for example, at a hub or whatever, like you know, end address requested by a customer, you just need to unhook that trailer and then free up that tractor for the for the next uh mission.

SPEAKER_02 4:11

And so for your your customer base right now, is there any active customer numbers or anything that you can share with us about you know, maybe some ROI that your customers are seeing you using this?

SPEAKER_00 4:24

So we are working with some of the um uh largest uh actually customers um in the country, you know, Ryan Transportation is a good example. And also we work with JB Haunt and and a bunch of leading shippers uh in the country. So uh of course, capacity is a major thing. We are uh experienced a significant shortage of safe drivers in the United States. And sometimes um, you know, even if you you uh you um want to pay people to do that and and you just can't find the people to do it, that's becoming a major issue. Uh, and it's gonna get worse uh in the next couple of years as well. So um, you know, the uh according to study, there's an estimate of uh 400,000 driver shortage uh in the country that's gonna happen in the next couple of years. The Department of Transportation, the federal government are uh onto that and the um, you know, they're aware of the shortage, and autonomous technology is the best solution uh for that. And also, you know, we got asked this question a lot. Like, are you know, are we working with customers where like even when there's no driver shortage at all, they still can't find labor to do that? And the answer is yes. A good example would be actually with dry transportation. Uh they have a hard time finding drivers to, you know, drive that that route at night on I-45 between uh Houston and Dallas. So um that the reason being like, you know, truck drivers do not want to operate at night. You know, everybody wants to uh have a good rest and they only want to drive uh during the day. But for autonomous systems, we love drive at night because it's more efficient, it's less traffic, we save a lot more fuel. And um, you know, this is actually a great opportunity for actually robots to pick up the job that humans don't do not want to do at all.

SPEAKER_02 6:17

So, what does I guess uh the onboarding process look like? Is there any onboarding process for the customers that you're bringing on?

SPEAKER_00 6:25

And that's what's really unique to our uh business model. The answer is no. We do not license our software, but instead we operate as a ship shipper. There's no onboarding whatsoever. It's the same as the customers. You know, we have our own DOT number, you know, we have the insurance, everything is the same as if you are hiring a human driver to deliver that that cargo. Everything is the same. There's no onboarding that uh that the customer has to do at all.

SPEAKER_02 6:54

Do the trucks still have to follow like an out-of-service rule, uh, even though there's no driver behind the wheel?

SPEAKER_00 7:01

We don't have to follow the 11-hour time limit. Um, so we can operate like 24 hours a day.

SPEAKER_02 7:07

Oh, interesting. And so from there, I imagine the the ROI is is is right there where you don't have to worry about a driver having you know getting too sleepy or falling sick or you know, something happening. Um they could just keep running or the the truck can keep running.

SPEAKER_00 7:21

Yeah, the robot never gets tired. You know, it's uh it's a tedious, it's a boring job. And many times, you know, when I drive, for example, on a road trip, um I sometimes get gets sleepy, right? And the robot doesn't do that.

SPEAKER_02 7:34

And so when you're delivering, are you delivering to the same facility over and over again? Or or is it different facilities?

SPEAKER_00 7:42

Uh that's a great question. You know, we do not build our own airports. I know, you know, some of the folks in the industry do, some of the companies do that, but but we don't do it. Uh, the reason being, like, you know, we don't want to create like operational problems for our customers. For example, like we don't want our customers to have to come to our warehouse and pick it up and then drive to their warehouse, right? That doesn't make any sense. That's additional cost for them. It doesn't really help our customer, it doesn't help the industry. And the reason that some companies do that is that they try to avoid as much as they can some of the challenging problems. For example, if they build a hub right next to I-45, then you can avoid almost completely uh driving on local road, right? But then, of course, like you know, your operational costs, it's gonna be extremely high. And for us, like, you know, we use public uh hubs uh sometimes, like you know, a rigged. You can rent a spot for only $10 a day. That's where, for example, Amazon trucks like, you know, uh pick up their cargoes, and sometimes you see a Walmart truck, or if you're in the state of Texas, you know, HDB truck as well. Um, so these are public spaces, these are not owned by us. We're just one out of many trucks that happens to kind of you know pick up our our cargo and drop off uh our trailer uh in these kind of spaces. Of course, you know, if the density between a, for example, two warehouses of a customer uh being very high, you know, we can actually very quickly open up that local driving. So instead of going to a public hub, we can go directly to the customer's warehouse if it economics work. And it only takes you know uh a couple of days, maximum weeks to open up like you know, uh a new lane. You know, it's not a problem for us.

SPEAKER_02 9:33

So what happens when the truck arrives at, say, you know, a warehouse or a DC and there's no driver to do a check-in? What does that process look like for pulling up actually to the warehouse and start unloading?

SPEAKER_00 9:47

So currently, like we still have operators within uh these hubs, the orb warehouses, to simply, you know, drive the truck around the warehouse, sign the paperwork, which is required by the law, and also like, you know, do perform the drop-up and hook and unhook operation. And also, you know, we we have like, you know, it's like flying an airplane, we have pre-flight checklists. So these operators also go through the checklist before they they send the trucks out. But in the future, you know, we uh we like to even minimize the operation within the warehouse as much as we can. Uh, for example, like you know, we can have the trucks automatically go like, you know, go into the parking spot, turn it off, and someone can, for example, just come over and unhook that and give the truck the gold signal. So um in the future we're gonna do a lot more autonomously, but right now the operation is very light in these warehouses compared to um you know the actual highway driving, which is like 95, more than 95% of uh operating the long-haul uh truck driving.

SPEAKER_02 10:53

Now there is a little bit of a debate around, I I think it was Elon Musk who who mentioned now how much he hates LiDAR, but LIDAR seems to be outfitted on the overwhelming majority of autonomous trucks. Uh what maybe is is is your philosophy around you know the types of cameras and the types of technology that are being used for autonomous trucking?

SPEAKER_00 11:12

We are practical. Like, you know, we use uh whatever sensors that we can find out there that make sense to use. There's a reason for LiDAR to exist uh because it has very good sensing of geometry and especially at near uh mid-range. So it provides a lot of safety redundancies for the entirety of the system. If you ask us, like, you know, can we operate without LiDARs? Of course we can. Uh, but we like to be even safer. The benchmark of Bob Auto's driver, virtual driver, is never human drivers because we are terrible drivers. We actually make a lot of mistakes and we have accidents. So our benchmark is how can we get 10 times safer than the human driver? And it's very easy to calculate, which is on the number of incidents per, let's say, like a million miles. And we want to have at least, you know, 10 times safer record, which means like only 10%, less than 10% of the number of accidents the human driver would make. And LIDAR at the current moment is instrumental in these uh for to achieve this kind of safety record. And the reason being that, well, sometimes we see these weird uh obstacle illusions, for example, with the camera only uh solution, right? Sometimes it's like, well, you know, we don't see actually there's anything there because of illumination, right? But when when you have a linear, you see that very, very clearly. And sometimes it's the other way around, like uh you see a shade, for example, from light, and somehow you feel like that's there's the obstacle right there, but actually, you know, um there's nothing over there, right? And uh we are practical on this because let's say um the cost of driving per mile right now is uh $2.7 with a human driver, and the amount of $2.7 per mile, $1.2 is related to a depression of the truck, as well as uh diesel, so so fuel cost. The $1.5 is human labor. So human labor is already over 50% of the cost per mile. And if we add a LIDAR, because our next gen uh built uh build material for the sensor rack is only $50,000 and it can be used for about a million miles. You know, it's it's not that expensive if we amortize these kind of fixed costs into the the whole life of the truck. So it's sort of a no-brainer to just pay a little bit of money, let's say 20, 30 cents uh per mile, uh, and we're already saving $1.5 per mile, right? And then we make the road safer, we make a lot less mistakes. That makes sense as a business model. If you're a a OEM, which means like you sell cars or you sell trucks, of course, you know, the the cost, the one time cost is very uh you're gonna be very sensitive to that one-time cost because your customers might be like, well, you know, if I go to another vendor, I can save $50,000 on that truck or you know, $10,000 on that passenger vehicle, you know, I want to save that money. But if we're operating this in transportation, the service business, whether we use our liner or not, is not making a huge impact on our cost per mile model. So, you know, we get an extra safety redundancy. And that's how we think of like, you know, what kind of things we want to use. Of course, maybe in 20 years, LIDAR is not important anymore. That's fine. We're just gonna remove them to save 20 cents out of MIR we try.

SPEAKER_02 14:44

When you're approaching new customers, do you do you find that there is, you know, uh, I guess a growing resentment towards using AI tools or you know, autonomous driving? There's uh, you know, obviously, you know, truck drivers feel a certain way about autonomous trucks. Um, but how are you combating some of the I guess the pushback or the challenges to adoption?

SPEAKER_00 15:07

That's an amazing question. I mean, before I joined bought auto and after I jumped out auto, I think about it every day. I've been working in um AI technology for more than 10 years. This is my 11th year in computer vision and robotics. And um I I did my uh PhD at UC Berkeley, and it's uh one of the best, if not the best, AI labs in the world. And I had a lot of different kinds of opportunities to work on AI, but I chose autonomous driving, but specifically autonomous trucks because robot trucks are must-haves in contrast to robot taxi, which I believe is good to have. We are having a severe driver shortage. So someone has to deliver all these packages for all of us, you know, for in 10 years, 20 years, because we're just running out of drivers. And this is an industry where just young people do not want to enter. Young people do not want to be long-haul truck drivers. Sometimes I hop on a truck and go to Dallas because for testing purpose, it's a tough job. Like you, you, you're in a car for you're in your vehicle for an entire day, you get very, very tired, and you're away from family, and that's only 200 miles. We're not even talking about 800 mile, 2,000 mile trips. It's a tough job. And that's why, you know, the average truck driver's age is already 46 or 47 in America. And the the age of people entering the workforce, this workforce is about 35. So it's not like after graduating high school, people become a truck driver. And people become a truck driver at a very late age, actually, the age of 35. And this age is getting higher and higher because the number of truck drivers retiring per year is significantly higher, actually, than like you know, young new truck drivers going into the market. So we're running on the people, right? Even if we kill it, let's say in five, six, seven years, we operate the largest fleet in the United States. That means we're operating 20,000 trucks. And guess how much cargo we're shipping for America about percentage point a year. About the percentage point. That's that that already makes us the biggest fleet of the country. So we're gonna have 400,000 driver shortage and we're gonna operate 20 to 40,000 trucks in five to seven years. That's only 10% of the labor shortage. And that's how challenging the problem is, right? When the shortage is this significant, this severe, we're not gonna disrupt the job market, right? We're we're not depressing the price. As a matter of fact, we never charge anything lower than the current market price because we don't we don't play that. There's no reason for us to do that. So imagine us as a supplement uh labor uh into the space, into a depleting labor force, and we're trying to do the best we can to sort of make this smooth transition. Now, how much American family on average is paying for, for example, classic trucks, the transportation by classic trucks? So the math is very simple, right? $500 billion a year in the industry, uh, about uh 130 million American households. So about you know, every household pays more than $3,000 on this because, well, you know, money doesn't appear on their own, right? So when the inflation hits, when the cost you when the price increases furthermore, it's gonna hit low to middle income family the most, which is 90-95%, right, of the entire country. And we're gonna save, try to save as much as we can for these folks. So, you know, this is unique to this kind of technology, to this kind of industry. It's because America's American economies build on trucks and we're running out of drivers. There's inflation has been a huge problem since the past couple of years, and low to middle-income families are hitting the most. So this is one of the very few technology that's um that can actually help the most society. It makes the roads safer. Um, it doesn't deplete, um, you know, it doesn't cause huge overnight disruption to a labor force. It's gonna be a smooth transition. And it's one of the most important challenges that we as a society in the country uh have to work together on. So, and that's my personal view. That's also the company's view. And that's why I chose to join this, right? We're not trying to replace a million, two million, five million uh white-collar jobs, which everybody likes to do. We're literally doing the things that humans don't want to do anymore.

SPEAKER_02 19:48

And so when you're you're talking about these long-haul routes, is there any opportunity in some of the last mile optimization, or is that still kind of uh a little bit far off for last mile versus that that first mile?

SPEAKER_00 20:02

Technologically, we're there. Uh, if we want to do like last mile, we can. Uh economically, it doesn't make as much sense uh because the driver shortage is mostly the long-haul driving inside the last mile. It makes sense because if my truck driver, well, you know, it's like driving an Uber versus driving a long haul, uh, let's say, like, you know, shuttle, right? Because then I get at least I get to go home every day. You know, I get to have dinner with my family, right? And it's more convenient. Actually, we don't see as much of a driver shortage for this last mile delivery or the type of labors that are closer to where people live than the long haul truck driving. So if we won't do it, we can, but we choose not to.

SPEAKER_02 20:44

That definitely makes sense, especially from the driver perspective on the lifestyle choices that they would prefer to make. And it's obviously more home time, staying closer to home, easier routes, things like that. Now, uh a final couple questions here is what do you see as sort of the I guess the outlook for the next couple of years? I know it's really difficult and really challenging to predict where AI is going, but but where do you where do you see AI in a couple of years relative to autonomous trucking?

SPEAKER_00 21:13

Well, um, a lot of people don't think in this way, but because of all the investment in AI and also uh huge breakthroughs in general AI and foundation models after 2022, well, that's the magic power of our technical stack. You know, we're the only autonomous trucking company that's founded after the AI revolution. That's why we're uniquely positioned to build everything right from uh from the very beginning, right? Our system is safer, it's more efficient, and everything. So we benefit a lot from these AI investment AI wave. In terms of predictability, I believe, you know, from the insider, I think it's perfectly predictable. You know, this year is a year of commercial viability. You're gonna see our trucks going nonstop every day from here. New Dallas and have it come back. And we're aiming for a positive growth margin by the end of the year. And that's going to be huge to the industry. Because for autonomo driving companies, even for leading robo taxi companies that are actually out there operating commercially, margin is still negative, meaning like, you know, people lose money when they drive around, right? And for us, like before we're going to scale, we're going to make sure our gross margin is positive. That just means for whatever trucks we buy, for how many trucks we operate, is always going to generate positive cash flow. So this year, we're going to achieve that. For the next year, it's going to be the year of initial scaling, scaling to many, many routes, scaling to states out of Texas. We're going to have more trucks because we have way more contracts than what we can convert actually on daily operations, because we're kind of strict on what time we should start scaling up or not. So the next year we're going to be the first year of scaling. And then the year, the two years after that, you're going to see us start to operate, for example, in New England to uh uh uh Michigan and all the way to uh Pacific West as well. So the year three and year four from now is gonna be like you know, dominating the industry and show that we are we're gonna become the largest fleet of the country.

SPEAKER_02 23:25

How do you choose which city to deploy next? Is it a particular kind of road or area of the country? Uh how do you choose you know, New England versus, I don't know, Tennessee?

SPEAKER_00 23:37

Uh most of that, of course, you know, is is done for an economic reason. Of course, you know, it has to default in our ODD, so our uh operational design domain. And the ODD is expanding significantly, right? For example, if we can handle snow driving, oh, we're definitely gonna go to Chicago, right? So that's expanding very significantly. But you know, we're a startup, and even if we're not a startup, you know, everybody has a limited checkbook. So we have to think of like you know where to deploy this capital and scale up the capacity. So we do have an actually in-house economist that does this calculation. Let's say like, you know, we have a million dollars that we can deploy. Uh, are we gonna deploy them on bombs on RD on trucks? Uh, and if you know we have to choose a lane, which lane is that gonna be? Is it gonna be a short haul lane or long haul lane or mid-haul lane between cities? And we actually calculate it based on like you know, density of the route, uh, if we are only going like one direction or bite uh by direction, is it gonna be seasonal demand or not? And all these things combined, actually. So, you know, we we can have an operational model for that. And the the strength of operating an autonomous fleet is that it can be seasonal, right? Sometimes when it's a season, for example, when there's a huge demand in Florida, and sometimes that only lasts for three or four months, right? If you're operating a truck company, it's mostly by uh driving by uh driven by humans, you have to convince a bunch of truck drivers to move from California to Florida for only a quarter and have them move back, right, in another quarter. And most of the time uh people cannot do that. But if we operate an autonomous fleet, we're able to um allocate resources based on forecasts of the demands, we can actually move, we'll move them around through the country, you know, and and and maximize our our return on capital.

SPEAKER_02 25:27

And then a last question here anything that you feel is important to mention that we haven't already talked about?

SPEAKER_00 25:33

Bot Otov is uh is a young company, but this is the company that's closest to solving the challenge. You know, but I think the industry needs to work together on that. Because, as I mentioned before, you know, once we operate the largest fleet of the country, you know, we're still only shipping like about a percentage point of the cargo. So all of us, all of the companies should work together on this. Now, there aren't a lot of companies right there working on autonomy trucks, because it feels like, you know, to a lot of people, logistics is a boring industry. But what we don't think of is like how much the entire economy depends on this. And it's like, well, if we close like a straight and somehow the world stops operating as we see today, right? Logistics is just exactly like that. Transportation is like that, trucking is like that. So it's a problem that all of us need to work together to solve and never view other companies as our competitors and view them as our partners. And I think the industry needs to work together with uh legislation, with investment, with uh infrastructure building to solve these challenges that all of us are going to face in the years to come.

SPEAKER_02 26:48

That's well said. I think it's a good place to leave the conversation. And where can folks follow you? Um, get connected with bot uh. auto yeah, bot not is it bot.auto?

SPEAKER_00 27:00

Bot auto.

SPEAKER_02 27:01

Bot auto. I think that was your web address, is what I was repeating there. Yeah, bot.auto is is the web address, and I'll make sure to include that in the show notes. But uh, where can folks follow you and get connected?

SPEAKER_00 27:12

Oh, well, you know, we are on LinkedIn um and we're on the places, so uh please follow us on LinkedIn and our social media.

SPEAKER_01 27:24

Thanks for tuning in to another episode of Everything Is Logistics where we talk all things supply chain for the thinkers in freight. If you like this episode, there's plenty more where that came from. Be sure to follow or subscribe on your favorite podcast app so you never miss a conversation. The show is also available in video format over on YouTube just by searching Everything is Logistics. And if you're working in freight logistics or supply chain marketing, check out my company Digital Dispatch. We help you build smarter websites and marketing systems that actually drive results, not just vanity metrics. Additionally, if you're trying to find the right freight tech tools or partners without getting buried in buzzwords, head on over to CargoRex.io where we're building the largest database of logistics services and solutions. All the links you need are in the show notes. I'll catch you in the next episode and go Jags.