Integrated Scheduling Of Order Batching And Delivery Under B2C E-commerce

B2C e-commerce,one kind of "Internet + retail" models,is quite popular with China's emerging middle class.You can shop anytime you want.Moreover you can shop various goods and enjoy the home delivery service.However with the big boom of B2C e-commerce,small lot-size,high frequency,and random dynamic arrival of customer orders offer big challenges to the order fulfillment.It is nessesary to use the order batching strategy to decrease the picking time.Thus how to pick up orders from tens of thousands of products and deliver them to customers in different locations in a shortest time or a lowest cost is a big issue in this work.As the strong application background of the order fulfillment problem,a lot of scholars study this area.Most of the studies focus on the order batching problem and vehicle routing problem,few study the integrated order batching and delivery problem.In the operation level,order batching process and vehicle routing process usually operate independently,and it is difficult to achieve the global optimum.Thus it is necessary to integrate order batching process and delivery process.In big data era highly shared information makes the integration of order batching and delivery possible.Based on the theory of the integrated production and delivery problem,we study the integrated order batching and delivery problem under B2C e-commerce.It has great theoretical significance and practical significance.As the problem is NP-hard.how to design efficient heuristic algorithm is also a key issue.There are two major delivery types under B2C E-commerce:the third-part logistics and self-built logistics.For example,Tmall.com uses the third-part logistics,JD.com and Amazon.com use the self-built logistics.Based on the different characteristics of orders under the third-part logistics and self-built logistics,we study four different types of the integrated order batching and delivery problems.Firstly we establish different mathematical models according to the characteristics of four problems.Secondly different algorithms are built to solve the models.Finally the effectiveness and efficiency of the algorithms are discussed by numerical experiments,the influence of different parameters on the algorithm performance is analyzed.In general,it is benifical to raise the order fulfilment efficiency by integrating the order batching process and order delivery process.It also provide effective decision support for the B2C e-commerce enterprise.The contributions of the study are summarized as follows:(1)The integrated on-line order batching and delivery problem without transportation time constraint under the third-part logistics.It is assumed that the distribution time is zero in the third-party logistics system.There are multiple pickers.The off-line version optimization model for the order batching and sequencing problem with multiple pickers is presented to minimize the turnover time.To solve the problem,the hybrid rule-based algorithm refers to the fixed time window batching rules is proposed in order to form batches and assign them to appropriate pickers under fluctuating order environment.Through a series of experiments,we discover several enlightening findings:1)the rule-based solutions demonstrate high efficiency in the turnover time with multiple pickers;2)the rule-based solutions lead to an impressive improvement of both saving times and wage costs under different arrival rates,picking devices and time intervals by comparing with the variable time window batching rules.(2)The integrated on-line order batching and delivery problem with fixed vehicle departure time constraint under the third-part logistics.We study the on-line order batching and sequencing problem where the vehicle departure times are known in the third-party logistics system.The off-line version optimization model for the order batching and sequencing problem with fixed departure time is presented to minimize the total service time and the total tardiness time.Several novel rule-based solutions are proposed,including the formation of batches,which are assigned to appropriate pickers without any information on the arrival times of future orders.The solutions define which orders are urgent and should be picked up directly,and which ones can be satisfied later.The solution algorithms are evaluated through a series of experiments.It is demonstrated that these algorithms can lead to a substantial increase of the number of delivered orders.Hiring more pickers is helpful to improve delivery rate and reduce waiting time,but the improvement range gradually decreases and reaches to zero.(3)The integrated on-line order batching and delivery problem under the self-built logistics.In the self-built logistics system,the orders are grouped to batches and delivered to the zone they belongs to by the capacitated vehicles.The round-trip time between the distribution center and each zone is constant.The off-line version optimization model for the problem is presented to minimize the sum of the makespan and the total delivery cost.We present an on-line 4-competitive algorithm by integrating the existing methods for the on-line integrated production-delivery on-line batching machine,on-line order batching,and find the algorithm robust and efficient according to the simulation.Through the comparison experiments with the benchmark,it is demonstrated that the choice of the algorithm can lead to a substantial reduction of both the total cost and the delivery cost.(4)The integrated off-line order batching and delivery problem under the self-built logistics.In the self-built logistics system,the information of orders is known beforehand.The orders are grouped to batches and delivered to the zone they belongs to by the capacitated vehicles with routing method.We propose a nonlinear mathematical model to minimize the time required to complete picking the orders,delivering to customer and returning to the distribution center,which solves the joint decision-making problem such as order picking sequence,picking process method and vehicle routing.For this NP-hard problem,a three-phase heuristic algorithm is designed.Firstly,the "clustering-vehicle routing" method is used to get delivery solutions.Secondly,the similarity-based order batching rules are used to optimize each route's orders.Thirdly,picking sequence is sorted based on the descending order of each route's delivery time.The experiments are proposed to test the efficiency of the model.The results are compared with the traditional optimization algorithm,which show that the three-phase algorithm can reduce the throughput time,decrease the vehicle's wait time and improve the delivery resource utilization.