Home Home Home

Didi Chuxing CTO Keynotes Symposium

March 29, 2018

Didi Chuxing CTO Bob Zhang on the Future of Automated Transportation

On March 7th, the U.S. DOT Center for Connected and Automated Transportation (CCAT) at the University of Michigan kicked off the 2018 Global Symposium on Connected and Automated Transportation and Infrastructure with a keynote presentation by Bob Zhang, CTO of Didi Chuxing. Zhang talked about DiDi's massive scale, its collaborative approach, and the company's ambitious plans for the future - highlighting many exciting implications for artificial intelligence and big data research.

DiDi launched as a taxi hailing app in 2012 and has since grown into the world's leading mobile transportation platform, providing a wide array of app-based transportation options for over 450 million users globally. "Currently we gather 70TB of new trajectory data per day, processing over 4500TB from more than 20 billion route planning requests," Zhang said.

This aggregated data is key to DiDi's vision for bringing better and safer transportation to cities around the world. "Our mission is to become a global leader in the transportation revolution through automotive technology and our strategy is AI for transportation," Zhang explained.  "In the past 20 years, the internet has solved the problem of information sharing. Well, we want to solve the problem moving of physical things. In the next ten years, we think technology will change transportation by optimizing transportation infrastructure, which includes smarter traffic lights and intelligent road network design. At the same time, intelligent driving technologies will greatly enhance safety-these are the roles of artificial intelligence, and possibly the most crucial challenges for us."

Engineers and data scientists at DiDi's research labs in Beijing and Silicon Valley are working with CCAT Director Dr. Henry Liu, Professor in Civil and Environmental Engineering at the University of Michigan, and researchers from other academic institutions, to share data and design future traffic systems. Zhang believes the future of transportation will be developed in three layers. The bottom layer involves the optimization of transportation infrastructure, which includes smarter traffic lights and intelligent road network design. Zhang explained that the company's smart transportation technology deployed in China can now predict demand and dispatch drivers to those locations 15 minutes before users actually hail their rides. The company has also helped implement smart traffic lights and other infrastructural changes in more than 20 cities in China, reducing congestion by as much as 20% in testing thus far.

A DiDi smart traffic light in China, which uses anonymized data from DiDi's platform to reduce congestion
Zhang continued, "the middle layer is the transition from gas-filled vehicles to new energy vehicles and the development of intelligent driving technologies. At the top layer, the relationship between people and vehicles will be redefined."

DiDi currently operates the largest electric car fleet in the world, with more than 260,000 EVs on its platform and plans to raise that number to more than 1 million by 2020.

Zhang discussed the value of partnerships for the transportation future, noting DiDi's long list of ride-hailing partners around the world, such as Grab, Ola, Lyft, Taxify, Careem, and 99, which it acquired early this year. "We have found that the technology and innovation we have accumulated in China can also help improve efficiency of transportation of other countries."
He also emphasized the importance of collaborative research projects, like those DiDi is working on with the University of Michigan, such as "Driver Performance Increasing through Team Competition," and "Deep Reinforcement Learning for Dispatching." Zhang said the University of Michigan researchers' work, in collaboration with DiDi, will "influence millions of people in the world," and "redefine the world of mobility for the better."