(12c) Advanced Available-to-Promise in Online Chemical Production Scheduling

In our fast-paced modern society, consumers have grown accustomed to the convenience of ordering goods online and receiving real-time information about availability and expected delivery date. This is widespread in the retail sector and is similarly gaining traction in the manufacturing industry. Offering competitive order promises and generating these offers quickly and reliably has become increasingly important for maintaining customer satisfaction and increasing market share [1, 2]. Order promising is a critical process in supply chain operations and resource planning. It involves determining whether a customer order can be fulfilled and, if so, assigning a reliable delivery date as well as scheduling production and distribution accordingly to ensure on-time delivery.

Available-to-promise (ATP) is a widely used framework in supply chain management systems for responding to customer order inquiries. It provides visibility into the uncommitted portions of the current and expected future inventory, enabling companies to respond to customer orders by determining the quantities to accept and generating accurate due dates [3]. However, most studies on ATP assume that the manufacturing plant has a constant capacity, ignoring any extended operational flexibility as well as variability in resource availability. This rather crude consideration of the manufacturing operations can lead to highly suboptimal or even infeasible order-promising solutions. To address these limitations, we apply an advanced available-to-promise (AATP) approach where we jointly optimize detailed order promising and production scheduling decisions. Although several previous studies highlight the benefits of AATP in discrete manufacturing, its application in chemical production remains largely unexplored.

We develop an AATP optimization model by incorporating order-promising decisions into the well-known state-task network (STN) scheduling framework [4], which effectively captures the complexities of chemical manufacturing. Additionally, we use a state-space modeling approach to ease the use of the model in an online scheduling setting [5, 6]. As customer orders arrive over time, they are collected within predefined batching intervals, and the AATP model is periodically executed to make order promising and production scheduling decisions based on the new information. Two computational case studies are conducted to evaluate the performance of the AATP model and compare it to the conventional ATP approach. Our results demonstrate that AATP achieves higher order acceptance rates, reduces inventory holding costs, and improves overall profitability. We show that while ATP only performs well when the demand forecast is relatively accurate, AATP’s performance is quite stable across different demand forecast scenarios, indicating its robustness under uncertain demand conditions. Our analysis further demonstrates that longer batching intervals lead to an increase in the number of rejected orders, as more orders fail to receive a response before the response due date specified by the customer.

References:

[1] Hosang Jung. An available-to-promise model considering customer priority and variance of penalty costs. The International Journal of Advanced Manufacturing Technology, 49:369–377, 2010.

[2] Christoph Kilger and Herbert Meyr. Demand fulfilment and atp. Supply chain management and advanced planning: concepts, models, software, and case studies, pages 177–194, 2015.

[3] Michael O Ball, Chien-Yu Chen, and Zhen-Ying Zhao. Available to promise. Handbook of quantitative supply chain analysis: modeling in the e-business era, pages 447–483, 2004.

[4] Emilia Kondili, Constantinos C Pantelides, and Roger WH Sargent. A general algorithm for short-term scheduling of batch operations—i. milp formulation. Computers & Chemical Engineering, 17(2):211–227, 1993.

[5] Dhruv Gupta and Christos T Maravelias. On deterministic online scheduling: Major considerations, paradoxes and remedies. Computers & Chemical Engineering, 94:312–330, 2016.

[6] Dhruv Gupta, Christos T Maravelias, and John M Wassick. From rescheduling to online scheduling. Chemical Engineering Research and Design, 116:83–97, 2016.