The Future of Transportation-Autonomous Trucks

Summary

In this episode of the Future of Transportation podcast, Julie Gates interviews Shahrukh Kazmi, Chief Product Officer at Volvo Autonomous Solutions. They discuss the future of autonomous freight, focusing on driverless trucks, safety principles, and the importance of redundancy in autonomous systems. Kazmi shares insights on Volvo's mission to address challenges in on-road trucking and confined segments like mining, emphasizing the need for safety and innovation. The conversation also touches on partnerships, real-world applications, and the benefits of driverless trucks in various industries.

Creators and Guests

Host

Julie Gates

Julie is the Chief Content Officer of The Transportation Channel and the Host of the TTC Podcast.

Producer

Matthew Perry

Matthew is the CEO of The Transportation Channel

Guest

Shahrukh Kazmi

Shahrukh is the Chief Product Officer of Volvo Autonomous Solutions.

What is The Future of Transportation?

This Podcast will focus on the future of transportation. Such as driverless cars, autonomous trucks, flying vehicles and much more!

Julie Gates (00:14)
Welcome to the Future of Transportation podcast with the Transportation Channel. Today we're starting a new series on the future of autonomous freight. We're talking about driverless trucks. I'm very excited today to talk to our guest, Shahrukh Kazmi, Chief

Officer at Volvo Autonomous Solutions. Hi Shahrukh, how are you?

Shahrukh Kazmi (00:33)
I'm good, Julie. Good to be here.

Julie Gates (00:35)
Good to have you. You have the job that I think a lot of young people dream about. You know, when adults ask little kids, what do want to be when you grow up and you get to play with trucks? How lucky are you?

Shahrukh Kazmi (00:46)
let me tell you this. I am the favorite at home because my son is super impressed with what I do. And I'm truly exciting job to do.

Julie Gates (00:53)
Yeah.

You're be cool for a few more years with your kids till they hit their teen years, right? Well, one thing I think is really interesting about you is some of your leadership philosophies because you're working in innovation and sometimes we need an innovation mindset when we do this work. So I know that you're a fan of Dr. Wayne Dyer and the Persian poet Rumi. How have they helped you set an innovation mindset?

Shahrukh Kazmi (00:59)
Exactly.

Yeah, I think it comes down to how you approach challenges in life and more down to how you take those challenges and make them into a growth kind of opportunities. That has really set myself in life and also at work when you are dealing with things which no one has done before, like an autonomous driving. So that really impresses me and it basically makes me run towards those challenges so you can address them. And it helps you grow.

Julie Gates (01:23)
Mm.

Mm-hmm.

instead of being.

Shahrukh Kazmi (01:47)
Yeah, exactly. I think some of the philosophy from Rumi and what Dr. Dyer talks about is just about ⁓ being confident with yourself and knowing that do your best and look for those challenges and grow.

Julie Gates (02:02)
Hard to do when you're in an industry where something's never been done before. You're really breaking new ground at Volvo.

Shahrukh Kazmi (02:07)
Exactly, exactly. And it's very hard to do. And I think it's this kind of mindset our team also has, where you have to look for challenges and issues so they can happen in our testing environments and so they don't turn up in the real world. So it really helps if you grow and move into that mindset, into those challenges.

Julie Gates (02:29)
That's exciting. Well, thank you for sharing that. I think we should start at the very beginning before we get into kind of all the safety, redundancy and the great work you're doing in that area. Let's start with Volvo Autonomous Solutions. What problems are you guys trying to solve?

Shahrukh Kazmi (02:36)
Thank

Yeah, very good. So here at Volvo Autonomous Solutions, our main mission is to bring autonomy to ⁓ currently into two big domains. One is on on-road highway hub to hub or on-road trucking. We are addressing that by being in US where we are ⁓ launching and doing commercial runs with our partners. And second one is looking into confined segment. One of those examples is like coal and mining segments.

And there in northern Norway and Brønnøy we are already running and addressing extreme weathering conditions where autonomy is really needed to address the challenges in those confined segments. And that's those two areas. And our core principle at heart is safety. And I think we're going to talk about it a few times when we go through.

Julie Gates (03:28)
Let's dive in. mean Volvo is known for safety. Didn't you invent the three-part seatbelt that has saved millions of lives? This is just infrastructure. It's who you are.

Shahrukh Kazmi (03:38)
Exactly, exactly. And while this thing is a true brand identifier, it gives us a lot of confidence, but it also really humbles us because it comes with a lot of responsibility. And that goes down into when we talk later on about why we approach the engineering and redundancies as we approach them, because we see and feel that responsibility which the brand brings, that we have to remain safe and we have to earn that trust.

Julie Gates (04:08)
So redundancy, explain how that applies to what you do.

Shahrukh Kazmi (04:11)
Yeah, so high levels. The concept goes like this. When you are working in complex and systems kind of related issues like autonomous trucking, ⁓ you need to have backups. You need to have backups because the human is not there. Human is an amazing component or machine in an ecosystem within that any failure happens has the ability to adapt to it. But when you remove the human, then we need backup systems like brakes, steering, so they can

fail in a graceful way. And that takes a lot of effort. You have to really think through what those concepts would be. And you have to embed that redundancy right at the core of how you build so you can scale properly because nothing else can replace human. So you have to go through a very systematic approach to decide how would I fail or degrade in a graceful way when something goes wrong, like breaking or steering or communication.

Julie Gates (05:08)
Let's go through some of those. I believe I heard you say there are six redundancies in particular you work on. Do want to cover some of those? Because I think the layman has no idea all of this thoughtfulness that Volvo is putting behind the safety of their driverless trucks.

Shahrukh Kazmi (05:12)
Correct.

Yeah

Exactly. And I think, let me start first with the redundancy is not new. Whenever you want to be safe, every time we fly in an aeroplane, there are redundant systems in there as well. We are between backup engines or multiple engines, communications. Redundancy is always there when you want to ensure high safety critical systems are in. And when you are on using, let's take the on-road example, autonomous trucking on highway, 40 tons on a truck,

on a highway comes with a lot of responsibility. So then you have to say, how do you make sure that it's safe and something goes wrong? It degrades in a graceful way. So we did a lot of focus work onto that to identify that we need to address in six redundancies. Those are braking, steering, communication, power and energy storage systems, computing. So this is where the brain of the software sits, which is deciding what to do. And then finally, vehicle motion control.

Julie Gates (05:53)
Yeah.

Shahrukh Kazmi (06:21)
which is all about the longitudinal and the lateral control of the vehicle. So it controls that.

Julie Gates (06:27)
Well, I noticed that you have more than one kind of, you don't just only use LIDAR. I now know that Volvo uses multiple ways of seeing, right? What are their systems? And those must be redundant in themselves.

Shahrukh Kazmi (06:40)
Exactly. so first I think is none of those sensors on its own can give us the functionality we need, which is to perform a function and be safe and reliable all the time. They all have weaknesses like cameras cannot look in has in low light conditions. LIDARs and radars have other challenges and adverse weather conditions. So you take all of them, fuse them, emerge them to build a very robust 360 view of the world around you.

But when you're looking into the computing, you also then plan different combination of sensors. So when if one set of sensors goes away, the other sensors can still help you to operate and perform. So you're building redundancy at so many layers to ensure that you can perform a function.

Julie Gates (07:25)
What about communication? That one kind of confounds me. What do you use to communicate when you're doing driverless trucking?

Shahrukh Kazmi (07:31)
Very good question. I like that because the communication is twofold. First, let's just look at the truck itself. The truck has lots of sensors, let's say cameras, radars, sliders. It has a computation module, but it has actuators like steering and braking, which has to get a message to move. If a network goes down in the truck, then the whole vehicle will not be able to operate. So that's like a single source of failure.

So what we then need to do is to put multiple communication channels. And we have to locate them very strategically within the architecture of the vehicle. So within an issue or an ⁓ accident or a slight nudge that when one of the cable is damaged, we have another route. So the network communication is never compromised. That's just one within the vehicle. But the autonomous driving is not just about the truck. It's about the ecosystem as well.

Julie Gates (07:58)
Right.

Shahrukh Kazmi (08:24)
So then you're talking about the infrastructure and the communication you have to do on the infrastructure. Then you have to talk about redundancy on 5G and how you're connecting to your infrastructure. So many layers there you have to add as well.

Julie Gates (08:33)
who satellites.

And I think I saw a great video you all have on your YouTube channel. So everyone go check it out. But you were talking about how you have different things on the front of the vehicle and the back of the vehicle. Is that because if there's something that happens in the front, it still operates because you have something in the back? What kind of redundancy is that system?

Shahrukh Kazmi (08:54)
Correct. think so the video you're referring to, so there are two. The one you're referring to, we are looking at confined segment. So in a confined segment like coaling and mining, there our operational domain, the environment we move in is very much controlled. So we need a lot more kind of insights from the front. And because we control the route at the back, we put different types of sensors. Whereas when you go on highway trucking, we put a lot of sensors on the front.

to build the picture higher up so it can look at the sensor itself. On the back, we put selected sensors to make certain use cases, like when something is coming from the back. But you don't need to replicate everything because sensors have very one good thing. If you position them right, at the right height, at the right location, it can give you a 360 view of the world. So you don't need to put them all the way, but there's still a few, which we put in.

Julie Gates (09:50)
What are some of the other redundancy categories we haven't covered? I know there were six. Which one's?

Shahrukh Kazmi (09:54)
Yeah,

so I think one which I didn't talk about and it doesn't come across, but the one I'm really proud of, very simple, is the one that when you put the trailer at the back, we also have the right at the back of the trailer, we have put in the lights, the brake lights. So what we do is behind the truck, if something goes wrong, we will have a redundant backup light. It's not a big thing, it's not very complex, but it shows our mindset to safety.

that if something goes wrong to the primary, just like a human would be driving, the brake lights, the secondary brake lights behind the truck are operating. And then we put another communication between the truck and the trailer, a standard trailer. ⁓ So if something goes wrong, we can still send the message behind the trailer. These are not called out as specific redundancy systems. We just do it because we see that we have to make sure that it is safe at all times and it should operate as if a human would notice that something is wrong.

And those are the things which makes me more proud because they show the philosophy of our teams.

Julie Gates (10:55)
Mm-hmm. Yeah, that's wonderful. ⁓ My understanding is there aren't any regulations in place yet for this kind of safety protocol. Is that true?

Shahrukh Kazmi (11:03)
Yeah, the regulation, the cool and the tough part of this job in the autonomous driving is that the regulations are not fully defined. ⁓ Where they are, they are slightly fragmented. They are in the definition of doing that, which means a lot of responsibilities coming to the OEMs or the builders and the virtual driver partners. And that's where we really have to raise the bar. So while the regulation of the redundancy is not there, we need to look for how do we make it state of the art?

How do we make sure that it is safe, true to our DNA and brand? And that's why when you look at that, we go back to the common principle of we will build redundancy. That's how we believe that redundancy is one of the key system, but it is not the only one ⁓ to make it safe. And I'm pretty sure as the industry evolves, it will see, and I have one example of aerospace industry, that it evolves and it becomes the norm that you know.

that in high critical system, you need backup systems and you need to be carefully deciding which those are. So I can see that the industry will evolve into a standard set of practices.

Julie Gates (12:11)
Hopefully you're leading the way through the work you're doing.

Shahrukh Kazmi (12:14)
Definitely, and super proud of my team for thinking in that mindset.

Julie Gates (12:18)
So another mindset that some other innovators have is speed to market versus redundancy and safety programs. You want to speak a little bit about to that and we can.

Shahrukh Kazmi (12:26)
Right, correct.

Yeah, you're 100 % right. when you are in advanced technologies like this, proving to the world that this, and we talked about the philosophy, visioning what the future may look like is so important. So speed to market has its value. But at the same time, if you want to build highly safety critical system, then they have to go back to the true brand of safety, reliable, uptime, all those things.

When those things are coming into the game, then you really have to take a step back and you have to decide that I have to build redundancy right at the start. And even if it slows you down, because it does, because with redundancy and you have to go back and look at architecture, it's not just as simple as putting two steerings and two braking systems in. You really have to look at your architecture. It gets very complex. It does slow you down, but it scales very fast.

once you get it right. So it becomes a scale enabler. And that's why our brand and our strategy and approach was to, right from the very start, build on a common architecture, shared technology, platform, concept, and approach, where we can build that from the ground up. So safety is not something you bolt on to show a function or autonomous driving. Safety is something you industrialize. And that's how you earn trust.

Julie Gates (13:50)
Very good. I know at VAS, the philosophy, one of the many, is partnership is the new leadership. So you've been in some interesting partnerships. Do you want to share what's going on with Aurora and the beautiful driverless trucks that are on the road as we speak? Can you share a little bit about that?

Shahrukh Kazmi (14:00)
Bye.

Exactly. So I think I'm going to start first with ⁓ how amazing and beautiful the truck looks. The Volvo VNL, every time I look at it, it's like very, very exciting ⁓ and it gives energy. Yeah. And then coming to your question, yes, in our confinement and quarry mining segment, we do our own virtual driver. We build that and we have gone, but when we went on highway segment, we decided

Julie Gates (14:21)
is interesting.

Shahrukh Kazmi (14:37)
that partnership is the new leadership, true to Volvo DNA and thinking from our leadership as well. And the approach there was to understand we, from an OEM perspective, bring scale uptime, how to build trucks. We have a very good network. We understand what to do. We understand the market. We understand the user experience. So we bring that. But if you need to move fast and move safely, we need to find partners who are aligned with our thinking of safety.

And one of those partners earlier on which we aligned with was Aurora. And I think a day or two, day on Sunday, if I'm not wrong, last week, Aurora themselves, I think very excitedly launched and did a run. I'm super excited for them. ⁓ And then ⁓ internally, we also have done a partnership with a second virtual driver company, very just recently, with Waabi ⁓ The concept here is that

we are building a scalable, safe and reliable architecture platform of a vehicle upon which we can integrate multiple virtual drivers. And that means Aurora and Waabi are already in public domain, which we have already aligned and we are working towards.

Julie Gates (15:52)
Okay, let's get like really detailed for our viewers who don't know what's going on in Texas. Can you explain what is happening there right now? This is groundbreaking.

Shahrukh Kazmi (15:57)
Yeah.

Exactly. in Texas, between Dallas and Houston, part of our team are currently running an autonomous truck with the safety driver. And I'll come back ⁓ why it is so important on the safety driver. Between Dallas and Houston, we are already running that with one of our partner DHL. And we are pulling real loads. ⁓ And we are learning. The mindset is about learning. ⁓

that we are learning from that experience, we are understanding what does maintenance mean, what would the feedback look like, and it is our team who is doing that. so autonomy is really on the road, and it is delivering real value to those customers. our customers are learning, and so are we learning, and our partners are learning. But that's not enough. I think our next one, we will soon be moving to another route within that domain from Fort Worth to El Paso. But these route selection...

or ODDs is very critical. And I will come back to that later on, that we have to be very careful how we decide which routes to go in so we understand that operational domain, so we don't change the weather too quickly. So in Dallas and Houston, we have our team which is running that, and we are learning at a very fast pace. And it is coming back into our engineering, and we are getting better and better with everything.

Julie Gates (17:24)
Okay, so there is a safety driver taking place in those threats. I was thinking they were just flying along by themselves right now, but you're still in the learning stages.

Shahrukh Kazmi (17:30)
Yeah. Yeah.

No, and then I want to say our technology is at the level that we are confident that it is ready for launch. But safety, I told you, also because of the brand, but also the mindset. Our approach is that we don't need the removal of safety driver is a very responsible step one should take. a lot of time has to while you are innovating at speed.

to develop something, one must slow down when it comes down to deciding, it time right now to remove the safety driver? And you only do that once you have understood and learned as much as possible. So you are building a good safety case and you're ensuring that all your requirements are met. And your data also tells you that, yes, now is the time. So we are running it with safety drivers. So the drivers are in place.

They also learn about the autonomous driving ⁓ and the technology. So they give us very valuable feedback. We need that because the price of ⁓ not being safe is too much. So it's really important to slow down and use that experience of our safety drivers back in the system and decide in a very data-driven way when it is the right time.

Julie Gates (18:48)
So this has been going on maybe a week, is that right? Did you guys launch the video as of the day? Yeah.

Shahrukh Kazmi (18:51)
For Volvo

autonomous solution with DHL, I think we did that a few months back. If I'm not wrong, it was December of last year, but I think it's a few months round. are already, with DHL, are already running. ⁓ And then I think our partners have done a few more runs in a different way, I think for a few days. But we're running, ⁓ for a...

Julie Gates (19:11)
Maybe it's to Houston that may

be the new one then. That might be what I was reading. Dallas to Houston route. That might be the new one that I'm just learning about myself.

Shahrukh Kazmi (19:15)
I say that again, sorry.

Exactly.

then even Dallas to Houston one we're running since, if I'm not wrong, since December without DHL party.

Julie Gates (19:25)
Okay, very good.

What are we learning? How are these vehicles working with, we've got regular passenger vehicles, they're public highways. What are we learning? How's the truck doing?

Shahrukh Kazmi (19:35)
Exactly. So we are learning at so many levels. Autonomous driving is like a team sport. So you're learning about the infrastructure, you're learning about the truck, you're also learning about when you are moving goods, what kind of challenges they will come. We are learning how will we do maintenance. ⁓ The vehicles perform differently with different use cases. They're running more often. So we're gathering all that data to understand, OK, what does that mean for maintenance?

our pre-checklist, what should they look like. And we are improving our processes with every day as we learn more and more from how the vehicle is behaving, how when they come back, we are doing the inspection things, then we are going back into the data to look how those components have behaved. And that data is getting back into engineering and they're getting more clever and sharper about ⁓ how to do that. But real thing, if I can add, we are learning our customer journey, because it is our

her team who is running it, for in that example for DHL, we are learning how the customer, our partners will experience it. And that is really a ⁓ true differentiator for us, because then we are understanding the true pain point of our customers and we're removing them as well from our processes. So it's an amazing place and it's an amazing team who is learning at so much levels to bring autonomy to the real world.

Julie Gates (20:48)
Yes.

It's very exciting. I want to take us maybe three steps backwards for people who are watching. Sometimes there's a little skepticism with just the general consumer who hasn't had exposure to all this great innovation and understanding what's going on. Can you explain what the benefits are to having driverless trucks?

Shahrukh Kazmi (21:18)
Well, let's start with just in US, between 2010 and 2050, the needs for transport will go five times. It will increase five times. We know we already have driver shortages. So this autonomy will not replace, but actually add. It will add value to that ecosystem. And with that adding value to the ecosystem, we will ensure that we can still deliver goods.

Julie Gates (21:34)
this.

Shahrukh Kazmi (21:47)
at a fast scale. People, for example, who wants to stay closer or the teams who want to stay closer to their home. It will improve the quality of their life as well, because you still need teams and people, but in a different kind of skill set. But the same kind of experience of autonomous trucking ⁓ is within close to our hubs, at our hubs, to monitor and maintain those trucks and actually manage those trucks.

And that experience to fill a real gap in the market, taking the US example, it's about not about replacing anything. It's about adding value where there is a significant gap. And it's a two differentiator for society. think the the vein of the society is about how the goods are moving, how the milk and all those things are turning up when when we need them. So for us, it's like it's a driving factor that it adds value to the society.

Julie Gates (22:43)
Also though, I think there are some interesting benefits when it comes to the work you're doing with mining and work like that. What are the benefits for driverless vehicles there?

Shahrukh Kazmi (22:50)
let me put it this way. Those sites are extreme in their weather. The example from Northern Norway in Brønnøy is the winter is very tough. And even the summer from our normal hemisphere standard is tough. So putting people all the time in that autonomous driving, which is a constant similar run in a very confined area.

We can improve not only the quality of life for those people, but also we remain safe. Those environments have high risk. The rocks are being blasted. There is rough terrain with which they're going through. Weather is changing extremely fast. By removing human from that element of danger and risk and putting an autonomous truck in that domain, which can run 24-7, it really adds value.

It reduces, it moves towards the vision zero of no accidents in such conditions. But it also adds value to the operators of those quarry and mining sites, because now they can run ⁓ at a much higher pace ⁓ continuously, ⁓ and they can still use the same team to deliver a far more bigger value while remaining safe, because one accident can bring a whole site down. And these autonomous truck.

then adds value by removing the driver from that unsafe situation or a risky situation.

Julie Gates (24:20)
also probably helps them with employment because aren't these sites usually really far away from where anyone lives and challenging to get to in the first place?

Shahrukh Kazmi (24:27)
Exactly, they're really far away, hard to travel to. And if you go to just I think we are operating in northern Norway, but some sites you even have to take specific dedicated planes just to get to those sites. So it really helps that human does not need to be there all the time and driving.

Julie Gates (24:30)
Right.

gosh, there's so much we could talk about. What other areas of redundancies and safety have we not covered today?

Shahrukh Kazmi (24:55)
Well, think the main one is this. There are three layers of innovation. One innovation is the autonomous technology. The second one is trucking itself. that's understanding trucking and being an innovator there. Volvo brings a lot of value. But the third one is the innovation in the safety of deciding when to remove the safety driver and being safe and then maintaining that process.

And that itself is a very innovative approach. Nobody has done that before. So how do you go and decide? There's no standard that you must follow. And if you have followed, there is a stamp approval that you're safe. So it comes down to us. So then you have to become a lot more robust in that process as well. And that's equally as hard as the engineering itself.

Julie Gates (25:43)
It's amazing, I could go on forever, but I think we wanna create a video series. Will you be willing to come back and do another show with us?

Shahrukh Kazmi (25:49)
100%. Yeah, let me know.

Julie Gates (25:52)
Great. That's Sharuk Kazmi, Chief Product Officer of Volvo Autonomous Solutions. Thank you for joining us for the Future of Transportation podcast. We'll bring him back and find out more about driverless trucks.

Shahrukh Kazmi (26:03)
Thank you, Julie.