Didi announced on Monday it is cooperating with the American education startup Udacity to hold an algorithm contest this May aiming to recruit talent on a worldwide scale to develop Didi’s competitiveness in machine learning and artificial intelligence (AI).
The president of Didi Research, He Xiaofei, said that contestants will have access to Didi’s latest transit data in several selected cities to develop an algorithm that predicts supply and demand for transportation. The champion will be offered USD 100,000, and the best performers will be invited to join Didi.
Didi Research, as the “brain” of Didi Chuxing, is now focused on projects including accurate ETA, dynamic pricing, car-pooling, predicting supply and demand and transport capacity dispatching, all being equally important, He told AllChinaTech.
He said that the projects have been integrated in the past three months, and the whole system is now working on defining dynamic prices, optimizing the matching of drivers and orders, predicting supply and demand and dispatching before the orders are placed. “Our goal is not merely to avoid traffic jams,” said He. “We work to prevent traffic jams.”
Didi’s advantage lies in its huge pool of transit data with 50TB of real-time data and 9 billion times’ route plans per day. Analyzing big data can facilitate machine learning, which in turn will help with the cloud computing that makes Didi’s intelligent transit decisions. With achievements made by Didi Research in the past year, Didi has decreased the ETA for pick-ups to the current four minutes on average and increased its rides to a total of 1.43 billion in 2015.
“Our vision for Didi is to grow into a company like Facebook and Google, a company featured by its own technological edge,” said He.
By cooperating with Udacity, who launched in China on Monday, Didi expects to welcome more talent to accelerate the realization of intelligent transit.
(Top photo screened from Didi Research)