Julian Ma, CEO @ Inceptio Technology: Commercialization of Autonomous Driving Technology

Updated: Mar 29, 2019

This episode features Julian Ma, CEO of Inceptio Technology, a LaaS (Logistics as a Service) company that operates a nationwide autonomous line-haul trucking network powered by their own autonomous technology. Julian is also the president of G7 Networks, an IoT start-up and the largest fleet management platform in Asia. Prior, he was the corporate VP at Tencent, the largest Chinese social media and gaming company. He oversaw the location-based services, search and autonomous driving businesses during his nine years at Tencent.

Julian shared his entrepreneurial journey and his view on the commercialization of autonomous driving technology through fleet trucking and logistic service.

Highlight 1: The math behind autonomous trucking

Highlight 2: Perspective on Truck Platooning

Watch the full interview:

Robin.ly is a content platform dedicated to helping engineers and researchers develop leadership, entrepreneurship, and AI insights to scale their impacts in the new tech era. Sign up with us to stay updated and access exclusive event, career, and business mentorship opportunities.

Full Transcript

Alex: Hello, this is Alex Ren, the founder and CEO of Robin.ly and TalentSeer. Welcome to Robin.ly AI Talks. Today, we're glad to have an honored guest Julian Ma, the founder and CEO of Inceptio. Hey Julian, welcome!

Julian Ma: Thank you. It's my pleasure to be here.

Alex: I know you founded this company together with some other partners. Tell us the story behind it: Why you guys founded this company and what kind of opportunities you saw?

Julian: Inceptio Technology was founded by strategic investors focusing on logistics technologies mainly in Asia and also by NIO Capital, which is a venture capital mainly focuses on transportation and new technologies. The reason that we started this company is that we see that in logistics industry in particular, in the interstate line-haul shipping, there is a real technology opportunity that we can use L3 technology to very much improve the operating efficiency of the fleet nowadays. So we see the opportunity is real, and also the opportunity is big because essentially we plan to be an autonomous tracking service platform in a major country. Our primary target today is China, but a few years later we are also targeting US market.

With that in mind, what we plan to do is that the company will have basically three parts.

The first part is autonomous driving development, this part of will be mainly done in Silicon Valley, since we still see Silicon Valley as the place, probably the only place to be able to develop a full stack autonomous driving system nowadays.

And we'll also have two other units in Shanghai. The second unit is what we call the vehicle engineering. So after the autonomous system has been developed in Silicon Valley, we plan to integrate our system into a few readily available line-haul trucks in China by working very closely with the OEMs in China. So for that purpose, we need a vehicle engineering unit in China.

And our third unit is actually the fleet operating unit. And this is going to be a big deal and a large unit for Inceptio Technology. We've already hired the best fleet managers and fleet operating professionals from the China logistics industries. So we will contract manufacturing those heavy duty trucks with our own technology from the OEMs in China, we are in advanced conversation stage with them to define who are our partners and the timeline for the production.

The current plan is that by the end of 2021, we will start the volume production of those L3 level autonomous trucks. And then we'll buy all of the trucks by ourselves, and we will run the fleet throughout China, mainly across the highway. So after we have a volume of those trucks, we will start with a few thousand trucks. And over the years, this will grow to maybe 50,000 or a hundred thousand trucks across highway in China. And with that, we'll run a sort of trucking service platform in China and we are offered the fleet manufacturing companies, logistics companies, the fleet trucking capacities on per mile basis. So that is what we intend to do.

Alex: It’s interesting that you chose L3 rather than L4. As far as I know, many other autonomous truck companies choose L4 in restricted areas, such as miles/docks operations. Why?

Julian: L3 on the highway is our choice of the market. The reason is that first of all, highway is a much, much larger market. On average, each interstate highway truck on an annual basis, their revenue on the per truck basis is 150,000 US dollars. And there are 6 million trucks in China. So 6 million times $150,000 per truck per year, that is actually close to 1 trillion US dollars. So this is a huge market. For example, even one percent of the market is enough to sustain a public company a few years down the road. So we see that for sure, this is a large market. And second, we see that with the L3 technology, which means highly automated, but still has a driver to take over whenever there is a need for the disengagement of the vehicle. The second reason is that technology wise, this is highly feasible.

Alex: In one of your previous talks, you mentioned that you guys not only focus on safety, but also on the improvement of operational efficiencies. You mentioned some numbers about 6 million trucks and what was the math behind it?

Julian: This is very interesting, because for the commercial vehicles, for example, the long-haul trucks, actually everybody's counting the numbers, so number is the king. The number is like this: on an annual basis, the cost for a heavy duty truck runs between 100, 000 to 150,000 US dollars. That is actually a fairly large number. Out of this number, roughly 25% of that goes to fuel, another 25 to 30% goes to drivers cost and actually the depreciation for the truck itself only accounts for around 5 to 10%.

With the L3 technology, we've seen that we will be able to reduce the truck drivers in Asia from two to one on per truck basis. The workload of the driver can be reduced by roughly 90%. So the driver will be reduced from a highly stressful driver into a very light job as an administrator. Because our goal is, for example, on average the truck will only needs a disengagement, which means to be taken over by the driver every other 30 minutes, the driver will take over, for example, for 30 seconds to one minute. And then they will return the control of the truck back to the computer. If we can reach that level of very low driver involvement, the driver’s job will become very standardized. That means that the safety level of the truck, the efficiency of the delivery of the job, and also the fuel consumption of the trucking, will not depend on specific drivers, which means that for example, for a journey from Shanghai or Beijing, in the past the job very much depends on specific drivers.

Because only this specific drivers know the truck, they know the road, and also know the truckload, they're driving the best. So that much ties up the driver to the journey to the type of the cargo and also to the specific driver. And so that means that one truck is very much tied to the two specific drivers that will carry a very high, for example, drivers cost. But when our technology is introduced to the market, that means any driver can sit on any truck for any journey at any point of time. So with that as a technology advancement, when we run a large network of the trucks, we can deploy any driver for any truck at the single time. For a single journey, we only need one driver. That would save roughly 12 to 15% of the labor costs on the project basis. This is a huge saving.

The second thing is basically since in most of the time, the machine will do the control of the vehicle based on road condition, based on truck load. So the fuel efficiency will always be kept at the best. That will give our customers another 5%, between 3 to 5% fuel efficiency. The depreciation for the equipment actually only accounts for 8 to 10% on an annual basis. So on the drivers, on the fuel, we will save roughly 15 to 20%, then we start from very small cost basis, which is 10% of the equipment. That means we can even double the acquisition costs, we can even double the price of the truck and the total costs will still be lower than today, will be actually much lower than today. So this is the calculation.

Alex: Truck platooning was one of the popular applications first introduced by Mercedes, and also some other companies, lately Volvo. What do you think about this application?

Julian: Actually, my view is quite different from generally favorable view on this, what we call truck platooning. Let me explain what truck platooning is. It means that two or three or even more trucks were around in a straight line. One track will run very closely to the other. The distance between the tracks is very close, it can be only like 5 or 10 meters, that is very close on the highway. Because only in this way, the first track will take over most of the resistance of the wind and the rest will just stay behind that the windshield. So on average, the trucks behind the first track will have the energy saving over roughly 10 to 15%. That is the sort of the foundation or basic assumption for the platooning. But that comes with a huge challenge with technology and also with the regulation. Because the platooning is running on the basis that the trucks will stay very close to each other. So one truck will only be, for example, 5 or 10 meters behind the track in front of it. I have a very hard time to assume that this very short distance will be allowed by any highway authority in the foreseeable future. So I think actually this requirement is even much higher than driverless for a truck on the highway for the safety issue and for the regulatory issue.

Alex: I remember you led a couple of investments in the autonomous driving industry such as Zoox. Only last week, Nuro, another autonomous driving company just raised $940 million. And Aurora raised about $500 million. Look at your investment and also what you do at Inceptio, what do you think about this market? I mean the autonomous driving industry, is it too hard?

Julian: I think it is not too hard, if you look at what autonomous driving will eventually bring to those commercial vehicles and the passenger vehicles. I just mentioned, there are three very big investments nowadays, Aurora, Nuro, even TuSimple yesterday. That is very interesting because if you look at Aurora, the value for Aurora, 50% of it will go to logistics and free shipping. Because the key sponsor for Aurora in the future is Amazon. That is probably one of the major reasons for the very high potential for Aurora. And for Nuro, it is very obvious, this is a 100% logistics sector, although it is in different subsets, but it is a logistics company. And for TuSimple, there's no need to explain, it is on highway.

That means that all of the big ticket investment, the big news as of this year are very much related to logistics, related to the free shipping commercial transportation. Because in this area, technology is much more feasible in terms of autonomous. By using this technology there is a possibility of establishing a very large, what we call it, a mobility service platform in this industry because there has never been a mobility service platform in this industry yet. Going in that direction, our next milestone of Inceptio will become nationwide free shipping robot network, that is even more exciting.

Alex: It sounds like automation is really the biggest opportunity for the logistics industry. But the logistics industry is a traditional industry. But Inceptio is a modern AI company. So when you guys try to disrupt this industry, what kinds of challenges are you facing?

Julian: Actually, we don't have that kind of view that this is a very traditional industry. I will say it is still a labor intensive industry. It's an industry that is highly dependent on technology and highly dependent on the financial leverage that is actually the driving forces for the logistics industry to move forward. In the US, it is already the fact that for example, Amazon is driving the technology of logistics. Or even say UPS, FedEx, they’re all investing in all kinds of technologies there. Autonomous driving is part of it. And actually, for example, they've already invested billions of dollars on the best freight aircrafts. So it is actually a high tech industry already.

For us, we see that over the last 5 to 10 years, the logistics industry has reached a stage that all of the traditional equipment, and all of the traditional ways of improving the human efficiency and improving the human flexibility have reached a limit. So the next stage to unlock the productivity bottleneck can only be through the sort of artificial intelligence in this industry. It is about autonomous driving, it is about robots. It can only by deploying a large number of robots because of the severe shortage of human labor in this industry. Because nowadays it has exactly the same situation in the US and in China. Nobody wants to be a truck driver. So the whole industry is facing a universal challenge of shortage of skilled labors.

Alex: Let's turn the topics to something about your career path. You started with electrical engineering during undergraduate, and you received a MBA degree from IMD. You have worked on management consultancy in A.T. Kearney. Those are all great experiences. How those different experiences shaped your view as an entrepreneur?

Julian: I never chose my career. I was pushed by others, or by the department of society or economy. In the 1990s, it was a time of telecommunication. So basically, I was looking for a job and was chosen by a telecom company. In early 2000, it was the time when China was booming, and there was a need for the management talent. So right after business school, I was chosen by a management consultancy, which is a very well known international company. In late 2000, it was the time that most of the internet companies in China were trying to scale up, they needed a stronger management talent. So actually, I was approached by them. But in any of them, I just enjoyed the work. I remember there was a speech by Steve Jobs at Stanford. It was about connecting dots. Jobs had done many types of jobs, particularly in early years. I found that all of the things as long as you enjoy, in the end they will pay back.

For example, now I am in an area which is a combination of autonomous driving and AI and logistics industry. But over the last 15 years, I also got involved a few times in the logistics. For example, when I was a consultant, I was helping Costco, which is the largest container shipping company to go public. That was 2001. And that was time that I started to learn intermodal, I started to get a first-hand experience with trucks. Before that I never had a closer look on trucks. So at that time, I just did that as a job, but I very much enjoyed it. I started trucking, I started container shipping and I also ran into the warehouses in the midnight, just to see how the workers move the cargos. So this kind of experiences actually now pay back, because I can understand by the first-hand experience, the value I can bring to this type of customers without doing the calculation, I can feel by heart. So I will say that, it's all about just enjoy your work, and apply a very high standard for any work I do. I stay many long hours, that's why I get so many gray hairs, but I just enjoy it.

Alex: I'm curious if you have made any tough decisions in your career?

Julian: Yes, there are a few. The most recent one is was in 2017 when I was still at Tencent. I was asked by the company to take on a troubling unit which was our search unit at that time. We had 2,000 colleagues working with us. It was very hard. But still we realized that there's no way we could catch up with the number one, because search and many other online businesses is a game of winner-takes-all. So we have to make a tough decision that shall we carry on like that way, still burning hundreds of millions of dollars on the yearly basis, or shall we find another way of doing that? It was a tough decision because first of all, that means that I wasted my time. It could even be perceived as my personal failures as a leader for the business, which is a quite large business and quite influential still in China. And it also means that how I could face our colleagues, they may need to choose another job, or they may even need to be merged or be arranged to some other businesses.

But still, if you look at what would be a sensible decision, we have to go back to the fundamentals - Does our business create enough customer value at that time? The answer is sort of marginal. Because by our way of doing things, we did not serve our customer better than our competitors. Does that make a commercial sense? No, because we are losing money every year and we're still No.3, No.4 on the market. So I had to prioritize all of the factors to make a sensible decision for the company and for the shareholders, and essentially for not wasting a few more years. For example, close to 2000 of our colleagues in that unit, although that would be very hurting to many people. So we did it, we merged our business with Sogou under the leadership of Wang xiaochuan, which is a very smart guy. In 2017, that business got listed on New York Stock Exchange.

Alex: For a lot of our audience who are watching this video, they will have a question when they are facing to be an experienced leader like you. So what's the most important trick for them to develop their leadership?

Julian: First of all, the leader has to have a big picture - already knows what the mega trend is, instead of being bothered or being dragged into the day to day details. Detail is very important, but the big picture and direction is very important because people are counting on you, basically your own future is counting on that. Second is about passion. Your passion will basically carry yourself as a leader to go far. Because at the end of the day, for example, for a successful leader, it is actually much less about money itself. It’s less about monetary or material benefits; it is more about making achievements. There are a few others. But I will say these are the two basic things that I feel quite identified with. And also, this is actually the things that when we are at Inceptio, choose or try to invite the leaders to join us. These are the things that we are looking for.

Alex: Specifically, I think that you mentioned you're looking for leaders. Can you give us more details about what kind of people you're looking for?

Julian: Our goal is to develop our site here to close 200 people. So with that in mind, actually we are hiring energetic, passionate, hardworking, curious engineers in all of these areas. We also very, very welcome all levels of senior scientists, engineering leaders to join us. It is going to be fun.

Alex: That's cool. It sounds like a very exciting journey to work at Inceptio. Thank you very much. This is Robin.ly AI talk and peer to peer conversation with Julian Ma. Thank you very much!

Julian: Thank you.