DiDi to Open its Big Data to Support Transportation Research

Author: Edison Mulia Oct 24, 2019 05:00 PM (GMT+8)

Through the collaboration with more than 20 cities, DiDi has helped improve traffic management in China, reducing traffic by 10%-20%.

Image credit: Markus Spiske on Unsplash

During this year’s China National Computer Congress (CNCC) summit held in Suzhou on Friday, Chief Technology Officer Zhang Bo announced that the Chinese travel platform giant DiDi (滴滴出行) would expand be opening its two extensive datasets, the Didigae Asia data and the Dripga Cover Asia data.

With the objective of easing the population movements, Zhang Bo explained the decision via a theory that the transportation sector will transform itself in three different phases. Initially, there would be an improvement of the transportation infrastructure, with the implementation of innovative intelligent technologies such as smart lights and road coordination. Next will come the automobile itself, where smart driving and electric cars will be the future trend. Last but not least, the most important stage will be the culture, as citizens shift more to new types of public transportation. However, the success of these different stages will rely on the involvement of AI as the core technology.

Not only is DiDi the world’s leading travel company, it has also extended to become a big data company. The ‘Drip’ platform currently processes an average of 4875 trillion data points, and receives 106 trillion vehicle trajectory data points daily.  Therefore, the staggering size of Didi’s dataset can be assumed to be a live map of transport in China. With the help of AI, this data can be extracted and converted to more valuable information, improving security, experience, and efficiency of travelling.

The ‘Outline for the Construction of a Powerful Country’ disclosed earlier this year by the CPC Central Committee and State Council, declared transportation as the national strategic priority. 

The corresponding data, which were taken from six major cities namely Suzhou, Haikou, Shenzhen, Chengdu, Xi’an and Jinan mainly concern trajectory information, road traffic index data and average travel speed. The rest contains two months’ of Didi’s Express and Premier service users’ travel information, as recorded in both Chengdu and Xi’an in late 2018.

DiDi has collaborated with more than 20 cities throughout China to improve their traffic management by implementing innovative smart transportation projects. This includes the optimization of traffic signals at more than 2,000 intersections in those cities, which resulted in the significant reduction of average congestion by approximately 10%-20%.

Through the application of big data and AI, Didi was able to also improve travel safety by broadcasting warnings to drivers such as when passing through accident-prone areas, or exceeding the speed limit, neutralizing avoidable accident.

DiDi released its Dripga Asian dataset earlier this year. More than 660 schools and research institutions have incorporated the data to their research.

It is hoped that with these data, better macro traffic analysis and forecasting can be achieved by combining traffic indicator and trajectory data. Zhang Bo further concluded that, in the future, DiDi will not only allow access to more datasets, but also continue to cooperate with academic institutions to further accelerate research development in the smart transportation sector.