Abstract

Meituan pioneered an on-demand food delivery service in China, which allows consumers to place food orders online for fast delivery. Leveraging its minute-level delivery network and boasting more than five million active couriers, Meituan delivers more than 60 million orders daily, ranking it as the world’s largest in terms of order volume and courier count. It has transformed itself into one of China’s most critical delivery infrastructure providers. To meet its service commitment to customers and improve the working experience of its couriers, Meituan must continuously optimize its order assignment decisions. Thus, it developed a real-time intelligent dispatch system, including a set of algorithms based on operations research and machine learning techniques, to precisely model the assignment problem in a dynamic and uncertain environment, and solved this NP-hard problem in seconds generating a high-quality solution. Since implementing this dispatch system in 2019, Meituan has realized a decrease of 20.96% in average order delivery time and 23.77% in average courier traveling distance per order. In addition, the system contributes to cost reductions of about $0.23 billion annually. Moreover, the dispatch system enabled other new digital economy business formats, such as Meituan Instashopping and Meituan Grocery, to thrive.

• Cited by
  • A high resolution gridded dataset for takeaway packaging waste in China

    10 April 2025 | Scientific Data, Vol. 12, No. 1
  • A Markov decision process framework for order dispatching in on-demand delivery services

    Multimodal Transportation, Vol. 122
  • O2O platform operations with self-scheduling delivery capacity