On The Eve Of Dada 618, The Technology Side Is Fully Upgraded And The Pressure Is Increased By 100%.

In order to effectively cope with the 618 major development of dada in the technology side ahead of a comprehensive upgrade: "Da Vinci" intelligent pportation capacity control system upgrade to join the intelligent preconditioning single function, on-line "big sky" big data platform, and at the platform level construction of AI algorithm platform, improve research and development efficiency, lay solid foundation for technology.

"Da Vinci" system upgrade again

On the eve of 618 this year, dada upgraded the "Da Vinci" intelligent pportation control system again, joined the artificial intelligence algorithm preconditioning single function, and achieved the order "side selection".

The "side picking and closing" function has been optimized from user orders, electricity supplier sorting, Knight orders and distribution.

As long as the order is placed, the system will automatically start the merge order logic.

We will make a comprehensive analysis of the receiving places, distribution distances and freight rates of other orders, and combine the appropriate orders such as "Shun Lu" with the consideration of Knight orders.

Unified to the store to pick up goods, pick up the goods and then send the Knights together to deliver.

After the function is added, the efficiency of order picking is expected to increase, and the average mileage of knights is saved by more than 30%.

Dada online "big sky" big data platform

According to Dada 618 project related person in charge: "during the 618 period of this year, dada distribution service covers 450 main cities in the whole country. It is estimated that the Jingdong's logistics order is about ten times more than that of daily growth.

As a result, the capacity of the dada platform will grow at least 4-5 times, and the participation of riders will reach millions.

So dada has launched the second weapon, the "big sky" big data platform.

It uses intelligent real-time computing technology, and uses trillions of information such as business data, spatiotemporal data and behavior data as the basic database. At the same time, it collects and analyzes the existing data.

It provides basic data support for intelligent order distribution system and power supply and demand regulation system. It also conducts real-time monitoring and analysis of store traffic, order, pportation capacity, commodities, performance and evaluation, and realizes Knight portrait, order evaluation, business feedback, business early warning and so on.

Through the "big sky" big data platform, dada has implemented the supervision of the operation of the knights in 450 cities throughout the country, and processed the large data and dynamically optimized the huge orders, deployed the pportation capacity to deal with unexpected situations, and ensured the stability and accuracy of the distribution.

The AI algorithm realizes high-speed wheel replacement.

On the eve of 618, dada launched the "full link pressure test" project, and conducted full link pressure testing on dada logistics distribution system on line.

Effective identification of the possible online risks of the platform system, and targeted optimization, provide a system guarantee for the 618 order peak scenario.

After full link pressure measurement, dada's capacity to cope with peak flow capacity is more than 100% higher than before.

It is understood that the whole link test project development cycle shortened by 50% compared with the past, human resources investment is only 50% of the original.

In the urgent need of business, dada not only enhanced R & D efficiency, but also reduced R & D investment and completed the "high speed wheel change" of high difficulty operation.

According to Dada technology department responsible person: "on the eve of 618, we further consolidate the technical capabilities of the platform, upgrading the construction of AI computing platform consisting of three major centers of computation, characteristics and models."

With the continuous expansion of business needs, the traditional technology development of business demand response will gradually appear "duplicate wheel".

This inevitably leads to the waste of R & D resources, resulting in the high cost of subsequent technical iteration and maintenance, which is not conducive to the precipitation and iterative innovation of the overall technical capability.

The AI algorithm effectively solves this problem.

Through the AI algorithm, the existing business can achieve faster iteration through the middle stage, and the new business can also easily reuse the existing component model.

The technology leader said: "through the AI algorithm, the R & D efficiency has increased by an average of 50%, saving a lot of development costs, shortening the development cycle and realizing the rapid expansion of business."