Below is the transcript of a CNBC interview with Henry Liu, VP & Chief Scientist on Smart Transportation, Didi and CNBC's Arjun Kharpal. The interview took place at CNBC's inaugural tech conference, East Tech West, in Nansha, Guangzhou.
AK: Thank you, CNBC, for giving me the opportunity to open, and kick off the first fireside chat here at East Tech West, and what better way to do it than talking about the future of transport and our smart cities. And Henry, you're going to give us an insight, because Didi is one-, or is the largest ride sharing service in the world. You guys do something like 30 million rides per day, which is amazingly big. So, what kind of data re you collecting, and how is that data being used to-, to inform your product?
HL: Yes, well, Arjun, before I start, I want to thank you for inviting me to be here, and to be in Guangzhou, it's such a nice city, it's a lovely city. It's a historical city, in Chinese history, and it's also an innovation hub for China also. So, talking about data, we collect lots of data every day. We serve 30 million trips per day, we collect 100 terabytes of vehicle trajectory data per day, so it's a large amount of data, and we process 480 terabytes of data also every day, so there's a lot of data being processed on our platform. So, how do we utilize this data? Well, we utilize this data to better serve you, as a passenger, so from both sides, and we utilize this data to-, when you open up our app, it will help you to locate, in terms of where you are, and we guess where you're going, and we find the best match for you, in terms of the vehicle supply, and then we estimate the price for you , in terms of how much you're going to pay for it, and we also provide a route for you, to your destination. So, that's how we utilize the data to serve you better. We also utilize this data to help the city better, and we work with the cities to look at, in terms of the traffic state, we evaluate what's going on with the traffic network, and we diagnose where the traffic problem is, and we help the city to optimize the transportation network, also. And, by the way, we can also predict, in terms of what's going to happen, in the next 15, 30 minutes, in terms of traffic flow.
HL: So, we use the data in a variety of ways, to help both our passengers, as well as the cities.
AK: So, when we talk about smart transportation and smart cities, these are two big buzzwords that have been thrown around a lot. How do you define that, when it comes to how you're using the data, and then how you then work with the governments?
HL: Right, so I think the most important thing for us is to use the data, as we just discussed, and we are a transportation company, not only-, we are not only an internet company, but we are also a transportation company. So, we use the data to help the passengers, as well as the cities, so-,
AK: And we've heard smart cities talked about, by not only you guys, at Didi, but many of the other internet companies here, in China, as well, and around the world. So, what makes your vision different to theirs, and how do you stand out?
HL: Right, I think there are a few differences. One is, as I mentioned, we are a transportation company, so we are offering a variety of products, and we have 14 different service products, including taxi, including express, premier car service, luxury car services, carpool, bike, e-bike, as well as food delivery. So, we are a transportation company, so we-, I think we have much deeper understanding, in terms of transportation content. So, that-, that's one. The second is that we work with the cities, we-, we are one of the users of the transportation network, so we work with the city, we complement with the public transportation system. So, for both of these, we are very different from other smart transportation solution providers.
AK: And how would you characterize where you're at, in the stage of using data to be more efficient in the way you allocate resource? Because I use the service here, in China, and I've used it in various cities, there is still a problem of supply and demand, where the wait times can be long. So, what are some of the learnings you've had, and where would you characterize , I guess, where you are, in the process of having this, kind of, 100% efficient transport network , with your service?
HL: So that's why we developed this product we call Smart Transportation Brain. So, essentially, it's a system level optimizer, to optimize the time/space resources. So, think about transportation. Transportation, essentially, is to serve the demand and also the supply. So, you-, for a transportation network, you have two-dimension network, plus time, so you have a time/space network, and you have people coming from different locations to travel on the network. So, essentially, you want to optimize how you utilize this three-dimensional time/space network. And when we develop this Smart Transportation Brain, we're trying to help the cities, and help our passengers, to allocate the resources, and use the resources more efficiently.
AK: But what have you learnt so far? Because it's not a perfect system yet-,
AK: So how-, how do you get to that state where there is zero wait time, or as near enough to zero wait time as possible?
HL: Yeah. So, there are, I think, a number of things we see there. We see the future transportation systems, with smart travel (inaudible) so that as long-, when you have a demand, you can utilize your app, to call for car services. And we also see, in the future, is ride sharing, right? So we don't have enough road space, for individual cars, and if we can share the ride, and if we can allocate the seats, instead of vehicles, so then we can use the transportation system more efficiently. So that's one. The second one is that the vehicles will become smarter, and so-, so we are also developing autonomous vehicle technologies, and that will help us in the future, in terms of the vehicle technology. And then, also, don't forget, we also have the city infrastructure, all of the vehicles have to be running on the infrastructure, so how we can manage our infrastructure more efficiently is very important.
AK: Yeah, I want to dig in to a few of those, Henry, because when we talk about smart cities, we often talk about driverless cars, as well, because we see these as an efficient way to run a city. Every technology company that's doing driverless cars has a different approach. Some are trying to do both hardware and software, some are just doing software. So, where is your-, what is your approach to driverless, at the moment, at Didi?
HL: So, at Didi, we develop autonomous vehicles from two fronts. One is based upon single vehicle intelligence, so we have the autonomous vehicle approach. Basically, we install sensors on our vehicles, we sense the environment, we detect objects, and then we do the route planning, and then we do the control, as well. So, that's on one side, on autonomous vehicles. The other side is what we call cooperative vehicle-highway systems, and it's more reliant upon the environment, the infrastructure, to provide the necessary information to the autonomous vehicles. For example, I'll give you an example of why this cooperative vehicle-highway system is very important. For example, if you have a big truck blocking you, the autonomous vehicles, and if you want to pass by, at that moment, if a pedestrian jumps out from in front of that big truck, and the individual vehicles will have limitation to sense that pedestrian walking in front, because it's being blocked by this big truck. However, if you have cooperative vehicle-highway system, and the highway system, the infrastructure can provide sensing capability, and that information can be transferred to-, transmitted to the vehicles, and the vehicle can react to it. So that's very important. So, to us, I think we develop from both fronts, from both single vehicle intelligence front, as well as from cooperative vehicle-highway system front.
AK: So when you talk about infrastructure, you're talking about sensors actually on roads, or perhaps buildings, lampposts, and the surrounding areas, in order to not just understand, you know, the car to car connectivity, but also the surrounding environment. Is that your approach?
HL: Yes, yes. So, the main difference is that we not only have the vehicle sensing capability, we're also going to have a roadside sensing capability, so we will be able to provide the autonomous vehicles with environment information, from the infrastructure side, and that obviously goes through either the LTE wave, or through the 5G capability
AK: Yeah, and I was going to ask, do we have the correct telecoms infrastructure at the moment to support such a vision? Because we have, of course, the LTE networks at the moment, the 5G is in the pipeline, but is 5G necessary for this to happen, just given the amount of data, and the fact that you need a low latency for this data, as well?
HL: I think so, I think 5G is very, very important, and in the US, we use dedicated short-range communication, as well. So, there are various communication capabilities, solutions available, and these solutions can provide the infrastructure for us to transmit the necessary information to the autonomous vehicles.
AK: How would you describe the stage of development you're at, when it comes to your vision for driverless cars at the moment? What are some of the steps you've done, and what are some of the achievements you've got, and where are you heading to next?
HL: In terms of our app?
AK: In terms of your own driverless car ambitions, and the whole infrastructure play, as well.
HL: Okay, so we have developed this autonomous vehicle technology, roughly three years ago, and we have development teams in both the US, as well as China. We have 40 autonomous vehicles being equipped with our sensors, and we have licenses in both Mountain View, California, as well as Beijing, China. So, we are on our way to develop these autonomous vehicles, and we also think we will be one of the frontrunners in terms of the autonomous vehicle technology development.
AK: And how do you see that being rolled out? Will it be small area by small area, eventually, until it gets across the whole of China? Or will it be large cities at a time? Where do you think you'll begin to see more mass scale of driverless car ride hailing functions?
HL: In terms of testing, I think we'll start with-, I think this will start with small areas, in terms of testing, but I think Didi has a lot of advantage in terms of autonomous vehicle development, because we have this transportation network, and we know where the origin, and where the destination is, and we know if this area is suitable for autonomous vehicle operation, as well. So, if it's suitable, we can send the autonomous vehicle, to serve for this trip, otherwise we can send a human-driven vehicle to serve it.
AK: Of course, you know, what we're going to hear about a lot at this East Tech West conference, is about responsible development of some of these new technologies, like AI, and, of course, like driverless cars. You know, this is a technology that has the potential to put a lot of people out of work, as well, so how do you deal with that disruption that could be caused from this new technology?
HL: That's a very good question. So, as I mentioned, we're not only an internet company, we're not only an AI company, we're also a transportation company, we transport people from origin to destination. That's a lot of responsibility, and we need to make this, safely and efficiently. So, we-, as a service provider, we utilize our-, we utilize our technology consciously to use this AI technology, and we also cooperate with the government. Our service actually complements with public transportation systems.
AK: And I want to-, as we close out this session, I want to get your views, you've spoken a lot about, of course, with driverless cars, what the future cities look like. If you were to give us your vision of, when this is all perfect, what does this look like for the consumer, how does this work?
HL: So, if you think about the future city, I think a future city will have much less in terms of parking spaces, road spaces, because we don't really need that much of spaces for vehicles. At that moment, I think we have an autonomous vehicle fleet, and they have-, they can serve the transportation demand, they don't have to park during the day, for doing nothing there. So-, and the road spaces can be smaller, as well. And so you will see in the cities a lot of green spaces, people can walk more, and utilize the spaces much more efficiently, I think.
AK: Henry, that was a fascinating insight, I'm excited to see what our future cities look like. Thank you so much for joining me today, Henry Liu, Vice President and Chief Scientist on Smart Transportation at Didi.
HL: Thank you.
CNBC is the leading global broadcaster of live business and financial news and information, reporting directly from the major financial markets around the globe with regional headquarters Singapore, Abu Dhabi, London and New York. The TV channel is available in more than 410 million homes worldwide.
CNBC.com is the preeminent financial news source on the web, featuring an unprecedented amount of video, real-time market analysis, web-exclusive live video and analytical financial tools.
CNBC is a division of NBCUniversal. For more information, visit www.cnbc.com.