The intelligent decision-making system is based on a large amount of data, using machine learning, deep learning and operational optimization techniques to establish a decision support system, which is applied to all aspects of production operations to help management and operators make more efficient and correct decisions. Improve efficiency, reduce costs, and control risks.

Take Suning Logistics as an example:

As early as the initial development of Suning Logistics, it was forward-looking to propose a long-term plan for the establishment of a new type of logistics, and the construction of an information management system was at an important position. After several times of polishing, Suning Logistics's self-developed intelligent decision-making system, the Tianji System, was officially put into use, like a “smart brain”, which promoted warehousing, transportation, and comprehensive digital and intelligent development.

The celestial system is put into artificial intelligence application in five aspects: forecasting, operation optimization, wind control, image and semantics. In terms of forecasting, machine learning technology is used to predict the sales volume of goods, the amount of operations in each link, and the timeliness of operations. Based on the forecast results, an optimization model is established at the planning level to further optimize the distribution of transportation networks and express points, thereby further optimizing supply chain decision-making and layout.

In terms of warehouse, Suning Logistics Tianji System obtains the optimal storage area of the warehouse and the optimal shelf type combination plan by establishing the operation optimization algorithm, and through a series of layout optimization such as commodity recommendation, inventory suggestion, replenishment plan and tally plan, The warehouse capacity utilization rate increased by 42%, and the comprehensive operation efficiency increased by 35%.

In terms of transportation distribution network, Suning Logistics has 2.3 billion annual transportation packages, more than 500 transfer distribution centers, and more than 4,000 trunk transportation lines. How to obtain the optimal route and reduce the entire network under the premise of meeting the timeliness Transportation costs are the core tasks of the Tianji system. Not only that, with the expansion of the smart retail ecosystem, the coordinated development of online and offline multi-formats also poses challenges for Suning Logistics' transportation network.

At present, there are 6,000 layouts of Suning stores, and the smallest unit in each city is the community. For Suning Logistics, the transfer business from the city center warehouse to the Suning store is from scratch, from small to many, and all the daily replenishment of the stores requires a complete supply chain system to support. The celestial system establishes an operational optimization model to solve the optimal strategy, rationally dispatch vehicle resources, and plan the distribution route; and can merge routes to reduce transportation costs, split routes to improve aging, reduce transit optimization costs, and ultimately help small store transportation network to save mileage. 34.7%.

In terms of end delivery, the delivery schedule helps the courier to properly plan the delivery sequence to shorten the delivery distance and improve delivery timeliness. It is worth mentioning that with the further integration of Suning Logistics and Tiantian Express outlets, as well as new scenes such as 3km 1 hour instant delivery around the offline stores such as Suning Store, the technical application of distribution scheduling is also constantly in Suning. Update iteration, the current version 3.0 uses dynamic programming + variable domain search algorithm, 8 single internal dynamic programming can guarantee the calculation result within 200 milliseconds, 8 single or more variable neighborhood search, guarantee the return of the calculation result within 500 milliseconds, 98% or more The single amount can achieve the optimal solution, and the daily delivery distance of the courier is directly reduced by 13.4%.

In addition, Suning intelligent decision-making system also applies image and video recognition technology to all aspects of logistics, establishes a wind control model, identifies abnormal behavior through video analysis, effectively manages operational risks, and makes abnormal operations invisible. At the same time, the use of voice and semantic recognition technology in the field of address management and intelligent customer service, greatly reducing labor costs.

Reprinted from the network