Yao Xinyu
OPTIMIZATION OF JD LOGISTICS INTELLIGENT SORTING SYSTEM BASED ON REAL-TIME DATA ANALYSIS *
Аннотация:
this paper examines the optimization of JD Logistics intelligent sorting system through real-time data analysis, exploring how cutting-edge technologies such as artificial intelligence (AI), Internet of Things (IoT), and big data analytics enhance sorting efficiency and accuracy. The study focuses on JDs innovative approaches to warehouse automation, route optimization, and predictive analytics, which have revolutionized Chinas logistics sector. By analyzing operational data from JDs smart warehouses and delivery networks, this research demonstrates how real-time data processing reduces sorting errors by 99.99%, improves operational efficiency by 40%, and enables same-day delivery for 92% of orders during peak periods. The findings highlight the transformative potential of intelligent sorting systems in achieving sustainable logistics operations while addressing challenges such as system integration, data quality, and workforce adaptation.
Ключевые слова:
intelligent logistics, real-time data analysis, sorting optimization, JD Logistics, warehouse, automation, predictive analytics
The history of JD Logistics intelligent sorting system reflects the rapid progress of Chinas logistics industry over the past decade. When the company first began building its logistics network in 2007, the sorting process was entirely dependent on manual labor. However, by 2025, JD had transformed its operations by implementing a multi-level automated system that includes autonomous vehicles, drones, and robotic warehouses [1]. The architecture of JDs modern intelligent sorting system consists of three interconnected modules: a sensor network, a real-time data analysis platform, and executive mechanisms. The sensor network includes thousands of IoT devices located in warehouses, vehicles, and even individual packages. These devices collect data on the location, weight, size, and contents of each item at a rate of up to 100 updates per second. The data is transmitted to JDs cloud platform, where machine learning algorithms process the information and make routing decisions. Of particular note is the automated guided vehicle (AGV) system developed by JD. These robots are capable of transporting entire racks to warehouse workers, eliminating the need to move them around huge warehouse spaces. Operational data has shown that the introduction of AGVs has reduced order fulfillment time by 30% and reduced picking errors by 45% [2]. Another key component of the architecture is JDs IoT plastic cargo bags with built-in sensors. These smart containers solve the problem of recognizing non-standard packaging, which previously made automatic sorting difficult. Thanks to IoT technology, the sorting accuracy of such bags has reached 99.99%, and operational efficiency has increased fivefold.Real-time data analysis has become the cornerstone of JD Logistics intelligent sorting system. The company has developed a comprehensive forecasting platform that combines three key services: internal inventory forecasting, key customer logistics order forecasting, and an intermediate logistics forecasting platform [3]. This system allows JD to allocate the necessary resources (labor, transportation) in advance, even during peak periods such as the annual Singles Day promotion. During Singles Day in 2024, JDs system processed a record 500 million orders, with 92% of first-tier orders delivered on the same or next day. This level of service was made possible by predictive analytics that take into account more than 200 factors, including historical sales data, weather conditions, traffic conditions, and even social trends [4]. JDs machine learning algorithms are continuously trained on new data, allowing the system to adapt to changing conditions in real time. JDs dynamic route planning system, which uses “digital twin” technology, is particularly impressive. During the COVID-19 crisis in Wuhan, this system simulated the impact of dozens of major road closures on hundreds of routes across the country in a matter of minutes and quickly adjusted logistics flows. Analysis showed that dynamic route optimization increased vehicle utilization by 40% and reduced delivery times by an average of 25%.Data quality remains a critical issue for real-time systems. JD has invested significant resources in creating infrastructure to clean and standardize data from diverse sources. According to the companys experts, the accuracy of forecasts directly depends on the quality of input data, and even a 1% error in the data can lead to a 15-20% deviation in forecasts.The integration of artificial intelligence and the Internet of Things has become a key factor in the success of JD Logistics intelligent sorting system. The company is actively implementing computer vision and deep learning to automate the identification and routing of goods. AI-based systems, such as GAINnext™ from TOMRA (JDs technology partner), are capable of recognizing objects in the same way as the human eye, allowing for the automation of sorting tasks that were previously performed manually [5]. Deep learning is particularly effective for solving complex sorting tasks that could not be automated with traditional optical equipment. For example, JDs system can now distinguish between food and non-food packaging based on their visual characteristics — a task that was considered impossible for machines until 2024. Training neural networks requires enormous amounts of data — JD experts upload millions of images into the system until it learns to distinguish the specific visual characteristics of material types, such as certain bottle cap or packaging shapes.JDs IoT technologies go beyond simple location tracking. Smart sensors built into warehouse equipment collect data on vibration, temperature, and other parameters, allowing the system to predict maintenance needs before breakdowns occur. According to JD, such predictive analytics has reduced equipment downtime by 30% and maintenance costs by 22% [6]. A distinctive feature of JDs approach is the combination of several sensor technologies in a single sorting system. For example, traditional optical systems based on near-infrared (NIR) and visual sensors (VIS) are complemented by deep learning technologies, enabling unprecedented sorting granularity. Now the system can sort not only by material type and color, but also by shape, size, dimensions, or other details. Route optimization is a complex multi-criteria task that JD successfully solves using a combination of operations research and big data analysis methods. The system takes into account more than 50 parameters, including current road congestion, weather conditions, order priority, vehicle weight and size restrictions, and delivery time requirements. JDs route optimization algorithms are based on the particle swarm optimization (PSO) method, which mimics the behavior of a flock of birds or a school of fish to find the optimal solution in a multidimensional space. PSO is particularly effective for logistics optimization tasks, as it can quickly adapt to changing conditions in real time. The implementation of PSO has enabled JD to reduce the total mileage of its vehicles by 18%, reduce fuel consumption by 22%, and reduce CO2 emissions by 15%.An interesting example is JDs smart delivery station in Yizhuang (Beijing), where the system analyzes data on more than 200,000 deliveries per month. JDs algorithms have determined that the optimal solution is for trucks to make two trips per day — in the morning and early afternoon — which allows the company to deliver on its promises of fast delivery at minimal cost. This approach, based on real data analysis rather than intuitive decisions, has been key to the efficiency of last-mile delivery.JD also uses “digital twin” technology to model and predict the performance of its entire logistics network. During the COVID-19 pandemic, this system made it possible to simulate the impact of road closures in Wuhan on hundreds of routes across the country in a matter of minutes and quickly adjust logistics flows. Analysis showed that the use of digital twins increased delivery reliability in crisis situations by 35%. However, the implementation of intelligent sorting systems faces a number of challenges. The main problems include the need to constantly update machine learning algorithms (every 3-6 months), dependence on the quality and completeness of data, and staff resistance to change. JD addresses these issues through continuous employee training and the creation of hybrid systems where people and machines work in symbiosis.
Номер журнала Вестник науки №6 (87) том 4
Ссылка для цитирования:
Yao Xinyu OPTIMIZATION OF JD LOGISTICS INTELLIGENT SORTING SYSTEM BASED ON REAL-TIME DATA ANALYSIS // Вестник науки №6 (87) том 4. С. 30 - 34. 2025 г. ISSN 2712-8849 // Электронный ресурс: https://www.вестник-науки.рф/article/24540 (дата обращения: 11.02.2026 г.)
Вестник науки © 2025. 16+