On November 29, it was announced that on the fourth anniversary of Didi's carpooling launch, Didi held the media open day to release new carpooling products, and announced that on December 3, Didi would launch an activity of carpooling in 26 cities with a discount of up to 100 yuan, to create the first 'carpooling day for all'.
Didi carpooling is a derivative product on the basis of express. The express drivers and vehicles provide services. The system matches the passengers with similar routes and shares each seat in the car, so as to share the fare and reduce the travel cost.
According to the data of didi platform, 2.9 billion people have used carpooling since the carpooling was launched, and 4.5 billion kilometers have been driven in the last year. About 30% of didi users in Beijing will choose to carpool during the peak commuting hours, which will help improve the urban traffic efficiency.
On the opening day, Sheng Kehua, general manager of didi car sharing, explained the concept of new car sharing in detail. It is essential for didi to create more value for users to promote the real carpooling. Through the refined planning and investment, the users can get more affordable taxi and more profits for drivers. So as to promote the sustainable development of car sharing and continuously improve the proportion of green travel in cities.
The two functions of the new carpooling can respectively meet the needs of users with timeliness and high cost performance. For users who pay more attention to fast departure, you can check 'accept carpooling' on the express page and then issue a ticket. Both the empty car and the passenger car that has received the carpooling can pick up and drive, so as to help users start faster when they are queuing for a taxi.
The "piecemeal departure" function enables users to meet the demand of taking a ride at a piecemeal price when starting. The user can send the order in advance on the independent page of carpooling, and the system will match the carpooling before calling. Placing an order in advance is convenient for the system to accumulate a certain number of car sharing needs, so as to improve the probability of assembly and the degree of going along the road.
For drivers, carpooling order is an important revenue component to reduce the empty driving rate and get more rewards. According to the data calculation, in the past month, the driver can reduce the empty driving rate by 16% when receiving the car sharing order. Compared with the express orders only, the driver gets about 14% more bonus for charging the express and carpooling orders at the same time. On the premise of not affecting the passenger billing, Didi is also exploring in 52 cities to provide drivers with multiple pricing of shared mileage, so as to make drivers earn more.
In order to maximize the utilization of seats in the express, first of all, there needs to be sufficient demand density. Second, the technical challenge is to predict the distribution of order supply and demand in the future period, and calculate the degree of traffic and the optimal matching of supply and demand. Zhang Bo, CTO of didi travel, said that at present, Didi uses the deep learning model to establish a real-time supply and demand prediction system, aiming to achieve the minimum detour, and at the same time, make the route matching of multiple passengers traveling along the road better.
At present, every time didi realizes the successful match between friends, it needs an average of 186000 additional matching calculations. Combined with more than 40 billion map path planning requests per day and the ability to process more than 4875tb of data per day, the average detour time of carpooling has been reduced by 30% compared with last year after continuous optimization of algorithm.
The realization of "sharing" means that it will consume time to pick up and send friends, and it will also be affected by unexpected traffic jams and other objective factors. In order to facilitate passengers to master the journey in advance, Didi added the prompt of estimated arrival time in the whole process of carpooling, and planned to launch the delay protection rights and interests. Through continuous technological innovation and experience optimization, we can achieve a better balance between affordable and convenient car sharing.