Continuing from an interview with founder and CEO of DiDi, Cheng Wei discusses his views on the future of urban transportation and the implications it may have for China. He also discusses important challenges and opportunities in the ride-sharing space.
New Trends in Urban Transportation
Ma Ping: DiDi Smart Transport has been leveraging existing data to improve transport. As far as I know, now you have abundant fund and talents, with the goal of reshaping the future of transportation. In 10 years, 20 years, and 30 years ahead, can you visualize the developments in transportation?
Cheng Wei: 30 years later, urban transportation will expand to three dimensions. Cars will fly, so brand-new traffic rules will have to be set up. As a result, transportation resources will increase by countless times to solve traffic jams, whereupon urbanization will take a great step forward.
Intercity traffic will evolve into hyper-speed railway three times as fast as sound. It will not be science fiction to see a 15-minute trip from Beijing to Shanghai. Hyperloop is now actually building a high-speed train 3 times of sound speed in Dubai. The train is not made of long cars as the trains we see nowadays, but of small cars linked with one another. At a circular train platform, 6 or 10 people get on to each car, which is later safely accelerated into the main track with the speed 3 times as fast as the speed of sound. The combination of hyper-speed railway and three-dimensional urban traffic will make real difference to our world.
If we only look 10 years ahead, there will be three major trends:
The first major trend is car-sharing. People will gradually realize that it is not the car but ride that they are asking for. If commuting can be tackled stably and effectively, it is not necessary at all for everyone to own a car. In fact, the use rate of private cars is only 5%. People spend 5% of their time driving while bearing 100% of the cost.
The 95% car idleness turns cities into huge parking lots, as many parking spots become even more expensive than cars themselves. Such cities see no prospect of sustainable development. Meanwhile, the inertia of a centuries-long automobile industry drives people to purchase even more cars.
In the past 20 years, every time the car population has increased by 20%, the road mileage increased only by 1% in response. This tremendous gap between supply and demand inevitably leads to traffic jams. Therefore, we need develop car-sharing, so that people use cars by demand, and do not necessarily own them when they don’t need them. Cars operate in the city during the day, and have their batteries charged in the suburb at night. In this way, the massive parking lots can be turned into parks and schools.
In most dense cities with fewer roads and less public transportation, car-sharing is more quickly becoming a reality. The cities in China and Brazil are good examples, while the progress of car-sharing will be slower in Japanese and European cities.
The second major trend is new energy, since the general amount of usage of new energy vehicles will surpass that of traditional vehicles consuming fossil fuels in 3 to 5 years. There is the global plan to stop the sale and use of traditional vehicles in 10 to 20 years.
In the past, people extracted petroleum and coal globally, and delivered them worldwide. In the future, we will construct several new energy power centers which are very large in scale and rely on wind or solar power. Then we will transmit cheap energy worldwide via ultra-high voltage network, so that most of the places will no longer need power plants.
In the past century, the contention for fossil fuels, especially for petroleum, has made the cities and our world become the way they are nowadays. The current situation will change following the promotion of new energy. The 1 million gas stations nationwide, for instance, will probably be replaced by 300 thousand charging stations.
The third major trend is the dawn of the age of auto-piloting and artificial intelligence. The automobile may become the first AI robots employed in large scale in the history of mankind.
In fact, humans are not really suitable for driving. The safe visual range of humans is only about 200 to 300 meters, while laser radar covers a much broader range. Humans only have an angle of view of 180 degrees, whereas machines can see 360 degrees. Humans can only rely on individual learning to gain experience, training them with mileage, but machines learn along with millions of other vehicles connecting to one “brain.”
Humans drive through complex roads by steering with their own experience, while machines can learn to take every turn as precisely as does Michael Schumacher. Humans control cars through senses and intuitions, whereas robots can control them so precisely on millimeter and even micrometer scale. Also, machines never get drunk or exhausted.
I believe self-driving vehicles will be much safer than human drivers. What does this increase in safety mean for humans? Every year, there are approximately 300,000 motor vehicle deaths in China, but in the future 90% of these may be avoided due to the increased safety of self-driving technology.
Of course, self-driving cars will also lead to unforeseen problems, including changes in traffic laws and new ethical dilemmas. However, if this technology can save so many lives, I feel there is no reason to hesitate in implementing it.
Self-driving technology is to cars like an operating system is to other technology. In smartphones and computers, there are only a few common operating systems, such as Windows and Linux for computers; and iOS and Android for smartphones. DiDi hopes that its own system will be widely used for self-driving technology.
Ma Ping: You just mentioned that cars will contain the first AI in human history. Has DiDi considered other changes to automotive technology? In the near future, we may not be able to advance to true AI, but is DiDi interested in developing a car with technology that surpasses modern vehicles? More generally, what kind of car do you envision?
Cheng Wei: DiDi is looking for strategic partnerships to design the first-generation shared vehicle, the “D ONE.” This car model, from day one, will be designed for ridesharing rather than personal ownership. Let’s forget about the 0 to 100 kmh acceleration and focus on safety, comfort, and operational efficiency. Germans invented automobiles, and then the Ford Model T popularized them. We are working hard to ensure that China is the birthplace of the first massively used shared vehicles.
Ma Ping: Apart from the shared vehicles that you’ve mentioned, what changes do you expect to see in public infrastructure, such as roads? Will DiDi participate in the planning and construction of new infrastructure?
Cheng Wei: Smart transportation is one of DiDi’s top priorities this year, and we hope to help intelligently design transportation infrastructure. We have installed more than 100 traffic signs across China, so that drivers can have access to real-time traffic data.
Traffic control devices include more complex problems. For instance, we are now designing a new type of reversible lane, which can adjust the travel direction according to traffic flow in real-time. A road with four lanes in two directions, for instance, can be adjusted to three lanes in one direction and one lane in the other during the morning rush hour, and can be reversed to the opposite during the evening rush hour.
Our first reversible lane is now operational in the city of Jinan, with real-time adjustable traffic lights. I hope the transportation infrastructure of the future can all be connected in an integrated network, where people, vehicles, and roads are all linked to a central controller in order to increase efficiency.
Ma Ping: Recent planning for the Xiong’An New Area has led to several new ideas for urban planning. For instance, homes can only be rented and are not for sale and investing in property is forbidden. Has DiDi Research thought about urban planning from start to finish? What kind of city would DiDi Research develop if given the opportunity?
Cheng Wei: We are participating in planning the smart transportation for the Xiong’an New Area. We hope to implement many new ideas, such as areas specifically for self-driving cars. We also hope to introduce new products, such as a “smart bus.” These are buses that will no longer need bus stops and fixed routes, but can use big data to determine the pick-up locations and routes by user demand.
Ma Ping: You are saying that there are no fixed bus stops, but instead a small bus could be dispatched whenever is needed to transport people?
Cheng Wei: Exactly. We can help commuters move more efficiently based on demand, and where they are. Before the Internet, commuters could not share locations with buses, so they had no choice but to go to a predetermined bus stop. With the Internet, there can be real-time and dynamic planning to optimize commuting.
Ma Ping: In Jinan and other cities, DiDi has been involved in basic transportation planning. However, most vehicles on the road are not DiDi vehicles. Does DiDi provide for ways other companies or individuals can make use of the systems DiDi is developing?
Cheng Wei: DiDi is still very far from taking over transportation planning. We hope the government will offer some technical guidance. DiDi will keep its systems open and hopes that other transportation systems that are developed are equally open.
The Price Mechanism
Ma Ping: DiDi services may be able to control the flow of traffic, but they cannot control supply and demand. Variable pricing is a tool that DiDi can use to do this. However, Chinese consumers are very sensitive to these price changes. As DiDi increases its role in the transportation market, how will DiDi deal with pressures from public opinion caused by variable pricing?
Cheng Wei: We hope that our efforts to introduce new products and technologies will make transportation easier for our users. We are promoting ridesharing and bus services to reduce cost and increase efficiency. In the future, ridesharing and renewable energy will also greatly reduce costs.
But pricing includes both economic and social factors. In addition to engineers, DiDi needs input from sociologists, economists, and psychologists. Determination of pricing is one of the distinctions between a platform and an enterprise.
Typically, a platform will not determine pricing. Prices are determined through supply and demand. However, an enterprise does make decisions on pricing. There are two ways to approach this issue. First is that DiDi is a platform, should not influence pricing decisions, and should allow users to negotiate directly with drivers. The second approach is that DiDi is an enterprise and needs to determine pricing.
Over the last few years, if you used DiDi premium service before you will notice that we did not do any price promotions, while the whole industry was doing that. We are figuring out ways to better serve the drivers as well, make them proud and esteemed, so that they will treat the customers with sincerity and smile. Then the customers will be satisfied by using the service.
Pricing is a complex issue, and one cannot talk about pricing without considering service quality. We will move gradually towards market-driven pricing, will keep our pricing mechanisms transparent to the public, and will give all price surge income to drivers.
I hope that in the future China will not just lead the world in transportation technology, but also in transportation services. We are looking to increase the income and social status of drivers in China while decreasing the cost and environmental impact of vehicles.
Taxi Services and Ridesharing
Ma Ping: Do you think the ridesharing services provided by DiDi will merge with existing taxi services?
Cheng Wei: That is a good question. Currently, DiDi has two main goals. The first is integration with taxi services, and the second is scaling across China. Near term goals include ridesharing and renewable energy. In the future, DiDi is focused on self-driving cars.
Integration of taxi services is a major trend for the ridesharing industry. At the start, there will be competition between taxi services and ridesharing services. Over the next two years, taxis will become more like ridesharing services in that they will need a system to assign rides and pools to drivers, and a rating system to measure the performance of drivers. Drivers who can provide better services to their passengers will earn more, based on these ratings.
Similar to what we have seen in New Retail, the line between online and offline ride-hailing services is being blurred. Taxi services will develop into ridesharing services, and as ridesharing services gradually become legal, they will change the face of taxi services.
We will test this integration in some cities by introducing our services to taxi drivers and allowing them to receive orders from our platform. In the future, if you order a ride from our platform and see a taxi coming to pick you up, don't be surprised! This will help the improve the service provided by taxi drivers, as technology is improving service.
Ma Ping: In some ways, taxi drivers can benefit from the model that DiDi provides due to increased efficiency; but in other ways, they lose the freedom to find customers. Do you think taxi drivers will accept this new model of mixing taxi services and ridesharing services?
Cheng Wei: By using a service like what DiDi provides, taxi drivers can be more efficient and earn more money. For passengers, if the only available vehicle is a taxi, why would you choose not to take it? Transportation is changing, and people will adapt to this new model.
Ma Ping: In routing transportation for an entire city, the software may no longer be optimized for an individual passenger. The self-driving car might not prioritize a passenger compared to the whole system, and some cars may be slowed down to increase overall efficiency. Do you foresee potential legal issues with this?
Cheng Wei: With limited resources, we must consider the benefits to the whole system. Ridesharing is a growing trend, and is an outgrowth of public transportation. Over time, people will stop buying personal vehicles and begin to utilize ridesharing services. The total number of vehicles on the road will decrease and parking lots will move further outside of the city. The percentage of people using ridesharing services will increase. Transportation will enter an era of self-driving vehicles, and it will be controlled precisely by algorithms and computers.
Consumers are searching for improved efficiency and convenience, which is what ridesharing provides. Three of the five largest privately held companies in the world focus on the sharing economy (Uber, DiDi, and Airbnb). These companies share idle resources to maximize efficiency.
In China, ridesharing makes up 80% of the sharing economy. How did this develop? Because when consumers have more power, it increases growth in all areas of the economy and transportation is the bottleneck to this growth. Ridesharing is the natural solution to transportation problems.
This article was originally published by Ma Ping via Guancha Syndicate on September 26, 2017.