The Direction of Autonomous Driving

Dan Feinberg | 04-12-2018

Flex (formerly Flextronics) is a company that I've worked with for decades, both as a customer and as a supplier. Recently, they were honored with Frost & Sullivan’s Manufacturing Leadership Award. I sat down with Eric Hoarau, the senior director managing automotive innovation at Flex, to discuss his company’s views on where autonomous driving technology is heading in the next 10 to 15 years.

Dan Feinberg: Eric, congratulations to Flex for being honored with the Frost & Sullivan Manufacturing Leadership Award. That was impressive.

Eric Hoarau: Thank you. It's part of our transformation. We are happy to see that the industry is recognizing some of the work we've been doing.

Feinberg: Let’s talk about autonomous driving and autonomous transportation. Where does Flex think this technology will be in five years? I mean, we're really at the beginning of it at this point. What do you see in the next five years or even in the next 10 years?

Hoarau: It's a good question, and one I think everyone in the industry is trying to answer—and everyone has a slightly different vision of where we're going. What we do know is that—especially for the next 10 years—it's going to be exciting to be in the industry. We see autonomous technologies advancing from two different directions: autonomy features and full autonomous vehicles. While autonomous vehicles capture everyone's imagination, autonomy features are what you're starting to see in all the commercials for cars, things like automatic braking or lane keep assist. Those features will be deployed at volume soon, and their purpose is primarily to help improve the safety of the vehicle and make people better drivers.

So the features that are going into mass-market cars will meet the Level 2 to 3 specifications on the 0−5 scale developed by the SAE. We’ve already heard the announcement that Ford was going to make things like automatic breaking  a standard feature in their cars. This is a big step for the industry and for safety on the road.

Feinberg: Very big step.

Hoarau: We are excited by that too. That's on the same order of things such as seat belts that improve safety across the board. In terms of what we call L4 systems, which are autonomous systems, those will come in different ways and forms around the world. The future is not just cars or pickup trucks, but there may be all sorts and types of vehicles and business models. It's going to be more service-oriented, and business models are going to go in many different directions.

If you have a high-end vehicle, you should be able to drive autonomously on the highway in four to five years, unless there are some major regulatory setbacks. Passenger fleets on specific roads and destinations in major urban areas will also be there. Again, it's not going to be the same everywhere, right?

You recently heard some interesting news about commercial trucks. That's also coming quickly, and the main reason is that they are bigger, more expensive, they can afford the equipment, and they tend to operate mostly on the freeway. It's a perfect fit for that. And then the last thing that people tend to miss sometimes is all the shuttle services that will be delivering goods.

Feinberg: I was going to say that's a big one. I saw some things at CES and in listening to the Nvidia people talking about their advancements in artificial intelligence and their contributions to this technology, I would think that the delivery of goods is one of the very first things we're going to see. You know, the van comes by your house every day at 10:00 and you pick out the groceries you need, and it automatically charges you, and it moves on its way to the next person.

Hoarau: Some version of that, yes. And it's not necessarily going to be a large vehicle. It really could be a small vehicle. People are getting used to it in some fashion. If you've needed to visit a hospital recently, they have robotic assistants that are delivering drugs and picking up things within a confined environment. You could argue that it's not fully autonomous and making independent-decisions, but it's starting to get there.

Feinberg: I agree. And I do think that that it is going to be one of the very first ones. You know, there's been some talk that there already are some autonomous or very nearly autonomous freighters moving across the world's oceans.

Hoarau: That's correct, and the same thing with airplanes, right?

Feinberg: Airplanes are a little scary.

Hoarau: But today, the technology is fully capable of taking off and landing the planes, but for regulations, the pilot does it. But once it's in the air, it's mainly controlled by the system.

Feinberg: It really is. Well one of the things I've been following is that in five or 10 years we'll soon see freeways that will be reserved for autonomous vehicles only, where a human being would not be allowed to control them. I think one of the biggest drivers of autonomous vehicles—and I'd like to have your opinion on this—is that most accidents are caused by human error. I think we're going to see a lot of arguments for autonomous driving in that it will reduce the number of accidents significantly. What's your opinion on that?

Hoarau: Yes. It's an interesting point. And, by the way, when we're thinking about being disruptive, the way we're looking at it is in a transformative way. Society is changing to hopefully bring more efficient productivity into the system and these technologies will help you with that. I like to think about it on the positive side and on how it can help us.

Feinberg: I absolutely do not think that disruptive means negative. I think it just means change.

Hoarau: Yes. And those are exciting changes, right? We look at it from the mobility perspective. Mobility is a broad space and autonomous vehicles are one component around that. And like you say, there could be lanes dedicated to autonomous vehicles, much like you have separate, dedicated bike lanes in many cities today. These are for safety but also to create new means of transportation that people can enjoy, and also help with congestion. I totally agree with you that this will be a great way for society to improve productivity, efficiency and most importantly, I believe, the quality of life.

Feinberg: I agree. When I sit on I-5 here in California to get someplace 20 miles away, and I've been in traffic for 40 minutes, I would love to have my vehicle driving me. I could do some work.

Hoarau: Yes, or relax. At least it would technically be safer, for sure. But from a point earlier, things like automatic breaking and lane assist in Level 2−3 cars are already starting to help with the safety aspect. With fully autonomous vehicles, it's going to be a sliding scale of improvement because it's not going to be the same everywhere, and not every vehicle will be autonomous. And, honestly, no one really knows yet what the future is going to look like. But for a fully autonomous system, you should expect the system to be better at avoiding human errors than what we have today. That's for sure. I think everyone agrees with that statement.

Feinberg: Let me ask you a few things about Flex's involvement with this. What segment do you think Flex will focus on?

Hoarau: We want to be the solutions provider. And that includes manufacturing and manufacturing at scale. Many systems today are small production trial and prototypes. We want to help companies scale those systems so that they can be deployed, both for full autonomy and what we like to call the journey to autonomy. We have a vision of full autonomy in our mind, but there are a lot of opportunities and things to be done along the way.

Feinberg: You're on the path and the path right now is taking us well into the semi-autonomous mode, but with the end goal being full autonomy.

Hoarau: Exactly. We see ourselves helping the manufacturer with many different parts to help them accelerate that transformation and build their products at scale. Flex is unique in this sense in that we play in so many industries at once, and this transformation is touching many of them. The autonomous vehicle is not just about the car or the vehicle, it's also about the cloud connectivity, it's about the servers on the back end. There are a lot of different pieces of the equation, and those are all areas that Flex is actively a player in today. We feel we can help by connecting the dots and helping them plan where the technology is going and see how we can get there.

Feinberg: It's also about autonomous vehicles communicating with each other.

Hoarau: You're correct. The vehicle-to-vehicle, the vehicle-to-infrastructure, that's coming up now in some form, but it would be there.

Feinberg: If you think about the tragic bridge collapse in Florida the other day, and let's say that there were autonomous vehicles. Well, obviously the bridge wasn't lying on the ground, so the autonomous vehicle wouldn't know about it, and maybe it would have seen it in time, but, with the communications now, one vehicle says to every other vehicle, “Hey, we've got a problem here, avoid this area.”

Hoarau: That’s right. What we've done in this space is to partner with a company called Savari, which is developing a software stack for vehicle infrastructure and then we're helping with the hardware piece of the equation.

Feinberg: Very good. Can you discuss what companies you're working with now?

Hoarau: Yeah, as you can tell, as the space is moving quickly; what I can say is that we're engaging with several customers at different parts of the stack. That means in the sensors, in the sensor fusion, in the autonomous computers, which is going to become more and more critical as you move to autonomous. An area that people tend to forget sometimes is the in-cabin experience. As we're moving to autonomous vehicles, it's not just about moving you from point A to point B safely, which is important, but also the user experience inside the vehicle is going to change, and understanding what technology needs to be there to help do that.

The companies we're working with are OEMs, as you could guess. The Tier One companies, some of the big names you probably know. And let's not forget the brand new companies, the ones coming up with unique approaches to the market and producing these new services. A critical part of this ecosystem is the new technology providers. Those companies providing things like the LIDAR solutions. We're working with them to help them scale their solutions. Those are pieces of equipment and systems that didn't exist five years ago.

Feinberg: And the leading brands, the leading companies that are announcing on this—we've seen announcements by Volvo, Volkswagen, and Ford, Tesla, and various parts of GM—I would say at this point probably a good answer is that there isn't anybody who's really involved in transportation that isn't looking at this at some level.

Hoarau: I think you're right. Like you mentioned earlier, we are at the beginning of the transformation. And that means there are many other players today which are looking at it, and there are the players which we don't even know about yet that will arise and will come up with new solutions.

But the important part, and you're starting to see in the news today if you look at the market, is that it will vary by region. Every region is going to have different players that will dominate because they can understand the localization of it, and it's a normal process in this space. I think the big discussion today that we're seeing in the industry is who is going to be in charge in the future? And there are a lot of articles and news about it, and people are debating it. Will it be the service providers? Will it be the car makers, or will it be entertainment companies? It all depends on the type of vehicles. To me, it's really fascinating, and I'm excited to be in this space. It's going to come down to where the new value is being generated because the pie is growing , and who is going to control that value generated?

Feinberg: It's interesting because of all the various disruptive technologies out there—and again I'm saying disruptive as change, not negative. Two of the ones that I'm most interested in, and that I'm following very closely—and it's interesting, I've spoken with Flex on both now—are of course autonomous driving and the other one is artificial reality, virtual reality, and mixed reality, and so forth. I've had some very interesting demonstrations and interviews recently and some of the stuff that's happening is just amazing what we're going to be seeing over the next 20 years in these areas.

Hoarau: You're correct, and I will argue at some point you can put those two together.

Feinberg: Oh absolutely, they will absolutely be coming together. I don't think there's any doubt about it. One additional question here: What do you think are the key advantages of autonomous driving? Let me give you some examples; there's safety, eventually there's cost. Right now, cost is higher, but I think ultimately it could be lower. There is ease of travel and reduction in labor. What do you see as the main drivers for autonomous transportation initially, and then maybe a little later?

Hoarau: I have a two-part answer on that one, and it comes back to the point earlier there are autonomy features and then there is fully autonomous. In terms of the autonomous features, it's going to be about safety and convenience, and that's a near-term driver for those technologies. This is what's going to put us at scale. This is going to be what brings money to the market to some degree.

When you're looking at the full autonomy, which is down the road, it's going to be about transportation-as-a-service, and every big player is looking around and saying, “Okay, what are the business models? How do we play in that space to be relevant in that future?” Everyone is different. There are many, many partnerships. If you look, there’s a complex web of partnerships of all the different companies across different industries trying to get to this future. And all that's going to be driving both the demands and the need for it and the financing for it.

As for the advantages, safety is there for sure because that's what society wants. We need it for safety. The way it is today, we need to improve it. Over the last two years, for the first time we have seen an increase in accidents on the road. And it's because of distractions, and many other things that we're doing in our lives. So those technologies should hopefully put us back on the right track towards reducing accidents.

Feinberg: There's no doubt that there are distractions while driving; you see people swerving around the road and, as you go to pass them, you notice that they're sitting there texting while they're driving.

Let me give you one other one that I've heard. It has nothing really to do with technology, but it's the use of real estate. For example, some very high percentage of vehicles that have been manufactured, are presently, at this moment, not going anywhere. They're parked. They're sitting there waiting for the owner, they're sitting there waiting for the company that needs them. With autonomous vehicles, you can have these vehicles running pretty much all the time, being used and therefore the need for a million and half parking places in every city is going to be somewhat less. It's kind of interesting how this is all going to start to come together and change society as we know it.

Hoarau: You're correct, and that's why it gets exciting for everyone to have this quality of life. Today you're using the car, and we must be fair to the car today, it's an extremely efficient and flexible way to travel whenever you want to go somewhere. And in that sense, it's an inexpensive way to give that flexibility and where you want to go. But in the future where we're moving, it's just going to be switching to those new services. That's where it moves from ownership to services. That includes everything from parking, to gas stations, to buying your groceries, to everything you do today. It will be exciting. It's not going to be a future which is just one way or the other. We really believe that it's going to be many different business models and use models which go in parallel and work together to get there.

Feinberg: Well, I appreciate you taking the time today. Is there anything that you would like our readers to know, anything else that you think that we should mention?

Hoarau: We talked about the fact that in the future the car is part of a larger ecosystem. We must work together, and we are, as a company, extremely excited to help accelerate that transformation at scale, because the benefit will mostly happen for everyone when we're going to scale, and at the right price.

Feinberg: Very good. Again, congratulations on that and we certainly will be following what you're doing with this.