Integrated lot-sizing and scheduling: Mitigation of uncertainty in demand and processing time by machine learning
Introduction
The integrated lot-sizing and scheduling problem (ILSP) is one of the most frequent problems in production planning that belongs to medium or long-term planning. In this problem, two types of decision variables are usually determined in a finite planning horizon, including master planning variables (the lot size, that is, the quantity of product to be produced in a given period) and the variables related to the detailed schedule (sequence of jobs within periods). The capacitated lot-sizing and scheduling problem (CLSP) with the resource capacity constraint is known to be NP-hard (Bitran and Yanasse, 1982). In addition, finding a feasible solution for CLSP is also NP-complete (Maes et al., 1991). This complexity intensifies in a dynamic environment so that meeting the client’s delivery time (without any shortage) is often challenging where initial plans are affected by unpredictable noise such as demand, processing time or lead time fluctuations, machine breakdowns, due date changes, order cancellations, etc. These disruptions generally trigger rescheduling and usually result in updating decision variables. Predictive, completely reactive, and mixed predictive-reactive strategies are three standard scheduling and rescheduling strategies for dealing with uncertainty. In predictive scheduling, all decision variables are made in advance, before realizing uncertain events, and no rescheduling takes place (Ouelhadj and Petrovic, 2009). Predictive scheduling requires a sufficient amount of information for prediction. However, the majority of real-world applications are indeed dynamic due to hard-to-predict events (Baykasoğlu and Ozsoydan, 2018). Completely reactive rescheduling is proposed for the problem, where all decision variables are taken in response to real-time events (Ouelhadj and Petrovic, 2009). The reactive rescheduling can be done either regularly after a specific time (periodic) or immediately after events are observed. The predictive-reactive strategy involves initial and predictive scheduling while all or some decisions regarding control variables are made in advance before uncertainty is observed. Then, in the reactive phase, all or some parts of the control variables are made or revised in response to real-time events (Gupta et al., 2016). Strengthening the schedule in the predictive phase reduces the cost of changes in the reactive phase. Improving forecasting and estimation capabilities in the predictive phase is crucial for this type of rescheduling. Hereupon, in recent years, the application of machine learning techniques in the form of data-driven optimization has been taken into attention. Data-driven optimization has been introduced as a new paradigm in optimization problems where evaluating the probability function or uncertainty is not straightforward. This approach emerged due to the explosion of data availability during the last decade and uses historical data derived from discrete event simulation or real production.
This paper proposes a predictive and periodic reactive strategy for rescheduling a dynamic CLSP in a job shop environment, while a data-driven approach using machine learning is proposed to tackle the demand and processing time uncertainty. We take a rolling horizon algorithm where scheduling is divided into periods, which means customer demand is updated periodically (weekly or monthly). This setup implies that even though we have information about upcoming periods, the more distant they are, the less certain we are about the uncertain parameters. Here we address this challenge by introducing safety reserves into the schedule. We consider two types of safety reserves: safety stock (a reserve in stock) and safety slack (leeway in the schedule). The question arises, how to set the values of these reserves? The values depend on external factors—for instance, seasons typically influence customer behavior. For this purpose, we employ use machine learning (ML) to predict values of the safety reserves for the next periods based on historical data.
Fig. 1 gives an overview of our approach, which is going through periods . The process might terminate early if the current demand or processing time does not enable a feasible schedule—this is undesirable, and that is what we are trying to prevent by setting appropriate reserves for the schedule. In each period, we first query the newly revealed data of the system (e.g., new customers’ demand) and then apply ML to predict the values of reserves. The existing information of the system and the values of reserves are then used to calculate the future schedule. For this purpose, we have incorporated a new model based on satisfiability modulo theories (SMT) in such a way that the model is calibrated using historical data and a custom-designed ML during the scheduling horizon. We use the solver Z3 as an exact optimization solver for solving the proposed SMT model. If the solver comes back with a feasible schedule, the current period is put into production, and the process moves on to the next one. If there is no feasible schedule, the whole process terminates. The algorithm’s performance is evaluated in a Monte Carlo simulation regarding the schedulability and optimality criteria as two critical measures in the face of uncertainty.
The rest of this article is structured as follows. Section 2 reviews the most relevant papers in the literature and the contribution of this paper. The description of the problem and the assumptions are defined in Section 3. Section 4 presents a novel SMT formulation of the problem. Section 5 illustrates the proposed solution method. Section 6 presents the computational results and Monte Carlo simulation to evaluate the proposed algorithm. Ultimately, Section 7 provides the conclusions and directions for future studies.
Section snippets
Literature review
Scheduling decisions are not integrated into the classical lot-sizing problems, and the usual approach, therefore, is to solve the lot-sizing problems first and then solve the scheduling for each period afterward (Drexl and Kimms, 1997). The discrete lot-sizing and scheduling problem, the continuous setup lot-sizing and scheduling problem, the proportional lot-sizing and scheduling problem, and the general lot-sizing and scheduling problem are some attempts to optimize the lot-sizing and
Problem description
We investigated the CLSP for job shop problems considering a sequence-dependent setup time. There are a predefined number of jobs (), each including several operations () arranged in a chain processed by capacitated machines with stochastic processing times. The scheduling horizon consists of periods (). The demand of job in period () should be satisfied at the end of period , while over the planning horizon, new random demands may arrive in the system (in
SMT deterministic formulation (model 1)
In this section, the SMT formulation of the CLSP is presented for a job shop problem. More details are as follows.
Indices j, j, k Index of jobs (j, j, k=1, …,J) o, o, l Index of operations Last operation of the job (j) m Index of machines (m=1,..,M) t Index of periods (t=1,..,T) Parameters Demand of job j at the end of period t Nominal value of processing time needed to produce one unit of product related to the operation Capacity of machine m during period t Setup time of if
Solution approach
There are several challenges in the rescheduling problem in different contexts. Ensuring schedulability in rescheduling is the main obstacle to deploying a new schedule in the face of variable demand and processing time. On the other hand, indicators of manufacturing performance, such as global objective function and schedulability, tend to deteriorate with frequent decision updates. To this effect, we propose safety stock and safety slack (2-SS) as a strategy to maintain the schedulability of
Computational results
In this section, the credibility of the presented SMT model and the performance of the proposed algorithm are evaluated and compared with other approaches. In this section, the following approaches are investigated.
- •
R-NN: The proposed algorithm that only uses neural networks.
- •
R-NN-Kmean: The proposed algorithm using neural networks integrated with the K-means clustering algorithm.
- •
Re-RH: The completely reactive rolling horizon where Model 1 runs at the end of each period according to the latest
Conclusions
This paper addresses the simultaneous optimization of capacitated lot-sizing and scheduling problems with non-permitted back-orders while demand and processing time are subject to uncertainty. A poor production plan causes companies cannot meet their customers’ demands, especially when the fluctuations of uncertain parameters are high. Therefore, schedulability is a prominent index in lot-sizing and scheduling problems with uncertainty. The current study proposes a new adjustable model
CRediT authorship contribution statement
Mohammad Rohaninejad: Conceptualization, Methodology, Software, Writing – original draft. Mikoláš Janota: Data curation, Investigation, Project administration, Funding acquisition. Zdeněk Hanzálek: Data curation, Supervision, Investigation.
Declaration of Competing Interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Mikolas Janota reports financial support was provided by Ministry of Education, Youth and Sports of the Czech Republic within the dedicated program ERC CZ under the project POSTMAN with reference LL1902.
Acknowledgments
The authors would like to thank the Editor-in-Chief and autonomous reviewers for their valuable comments, which greatly improved the quality of this manuscript. This research has been supported by the Ministry of Education, Youth and Sports within the dedicated program ERC CZ under the project POSTMAN with reference LL1902.
References (60)
- et al.
The multi-item capacitated lot-sizing problem with safety stocks and demand shortage costs
Comput. Oper. Res.
(2009) - et al.
Dynamic lot sizing with stochastic demand timing
European J. Oper. Res.
(2022) - et al.
A computational study of the general lot-sizing and scheduling model under demand uncertainty via robust and stochastic approaches
Comput. Oper. Res.
(2018) - et al.
Approaches for the joint resolution of lot-sizing and scheduling with infeasibilities occurrences
Comput. Ind. Eng.
(2021) - et al.
Dynamic scheduling of parallel heat treatment furnaces: A case study at a manufacturing system
J. Manuf. Syst.
(2018) - et al.
Scenario-based planning for lot-sizing and scheduling with uncertain processing times
Int. J. Prod. Econ.
(2006) - et al.
A new method for robustness in rolling horizon planning
Int. J. Prod. Econ.
(2013) - et al.
Heuristic-based neural networks for stochastic dynamic lot sizing problem
Appl. Soft Comput.
(2013) - et al.
Adaptation and approximate strategies for solving the lot-sizing and scheduling problem under multistage demand uncertainty
Int. J. Prod. Econ.
(2018) - et al.
A machine learning-based system for berth scheduling at bulk terminals
Expert Syst. Appl.
(2017)