Evaluation of heuristics for scheduling in a flowshop: a case study

The problem of finding the best permutation schedule for a flowshop has engaged the attention of researchers for almost four decades. In view of its NP-completeness, the problem is not amenable to the development of efficient optimizing algorithms. A number of heuristics have been proposed, most of which have been evaluated by using randomly generated problems for a single measure of performance. However, real-life problems often have more than one objective. This paper discusses a live flowshop problem that has the twin objectives of minimizing the production run-time as well as the total flowtime of jobs. Five heuristics are evaluated in this study and some interesting findings are reported.     

An efficient approach for large scale graph partitioning

In this paper, we consider the problem of partitioning the set of vertices of a graph intok subsets of approximately the same cardinality in such a way that the number of edges whose endpoints are in different subsets is minimized. A greedy heuristic, called Procedure1, based on the independent growth of each subset by fronts is proposed for constructing a good-quality graph partition. In addition, we present a more sophisticated version of the greedy heuristic, called Procedure2, which periodically calls a routine to refine the partition being built. We show that the partitions generated by Procedure1 are competitive with those obtained by several constructive heuristics in the literature, e.g. spectral, geometric, as well as other greedy heuristics. Moreover, the partitions produced by Procedure2 are compared with those produced by a leading graph partitioning method in the literature.     

Locating independent facilities with maximum weight: Greedy heuristics

The problem is to locate a maximum-weight set of facilities such that no two are closer than a given distance from each other. The unweighted version is equivalent to the maximum independent set problem in graph theory. This paper presents four greedy heuristics and shows that they all have bad worst-case behavior. Empirically, however, these heuristics perform quite well in the relatively large test problems generated randomly.     

Effective heuristics for the blocking flowshop scheduling problem with makespan minimization

The blocking flowshop scheduling problem with makespan criterion has important applications in a variety of industrial systems. Heuristics that explore specific characteristics of the problem are essential for many practical systems to find good solutions with limited computational effort. This paper first presents two simple constructive heuristics, namely weighted profile fitting (wPF) and PW, based on the profile fitting (PF) approach of McCormick et al. [Sequencing in an assembly line with blocking to minimize cycle time. Operations Research 1989;37:925-36] and the characteristics of the problem. Then, three improved constructive heuristics, called PF-NEH, wPF-NEH, and PW-NEH, are proposed by combining the PF, wPF, and PW with the enumeration procedure of the Nawaz-Enscore-Ham (NEH) heuristic [A heuristic algorithm for the m-machine, n-job flow shop sequencing problem. OMEGA-International Journal of Management Science 1983;11:91-5] in an effective way. Thirdly, three composite heuristics i.e., PF-NEH_LS, wPF-NEH_LS, and PW-NEH_LS, are developed by using the insertion-based local search method to improve the solutions generated by the constructive heuristics. Computational simulations and comparisons are carried out based on the well-known flowshop benchmarks of Taillard [Benchmarks for basic scheduling problems. European Journal of Operation Research 1993;64:278-85] that are considered as blocking flowshop instances. The results show that the presented constructive heuristics perform significantly better than the existing ones, and the proposed composite heuristics further improve the presented constructive heuristics by a considerable margin. In addition, 17 new best-known solutions for Taillard benchmarks with large scale are found by the presented heuristics.     

New high performing heuristics for minimizing makespan in permutation flowshops

The well-known NEH heuristic from Nawaz, Enscore and Ham proposed in 1983 has been recognized as the highest performing method for the permutation flowshop scheduling problem under the makespan minimization criterion. This performance lead is maintained even today when compared against contemporary and more complex heuristics as shown in recent studies. In this paper we show five new methods that outperform NEH as supported by careful statistical analyses using the well-known instances of Taillard. The proposed methods try to counter the excessive greediness of NEH by carrying out re-insertions of already inserted jobs at some points in the construction of the solution. The five proposed heuristics range from extensions that are slightly slower than NEH in most tested instances to more comprehensive methods based on local search that yield excellent results at the expense of some added computational time. Additionally, NEH has been profusely used in the flowshop scheduling literature as a seed sequence in high performing metaheuristics. We demonstrate that using some of our proposed heuristics as seeds yields better final results in comparison.     

On the Robust Single Machine Scheduling Problem

The single machine scheduling problem with sum of completion times criterion (SS) can be solved easily by the Shortest Processing Time (SPT) rule. In the case of significant uncertainty of the processing times, a robustness approach is appropriate. In this paper, we show that the robust version of the (SS) problem is NP-complete even for very restricted cases. We present an algorithm for finding optimal solutions for the robust (SS) problem using dynamic programming. We also provide two polynomial time heuristics and demonstrate their effectiveness.     

A novel discrete artificial bee colony algorithm for the hybrid flowshop scheduling problem with makespan minimisation

The hybrid flowshop scheduling (HFS) problem with the objective of minimising the makespan has important applications in a variety of industrial systems. This paper presents an effective discrete artificial bee colony (DABC) algorithm that has a hybrid representation and a combination of forward decoding and backward decoding methods for solving the problem. Based on the dispatching rules, the well-known NEH heuristic, and the two decoding methods, we first provide a total of 24 heuristics. Next, an initial population is generated with a high level of quality and diversity based on the presented heuristics. A new control parameter is introduced to conduct the search of employed bees and onlooker bees with the intention of balancing the global exploration and local exploitation, and an enhanced strategy is proposed for the scout bee phase to prevent the algorithm from searching in poor regions of the solution space. A problem-specific local refinement procedure is developed to search for solution space that is unexplored by the honey bees. Afterward, the parameters and operators of the proposed DABC are calibrated by means of a design of experiments approach. Finally, a comparative evaluation is conducted, with the best performing algorithms presented for the HFS problem under consideration, and with adaptations of some state-of-the-art metaheuristics that were originally designed for other HFS problems. The results show that the proposed DABC performs much better than the other algorithms in solving the HFS problem with the makespan criterion.     

Subjective machines: Probabilistic risk assessment based on deep learning of soft information

For several years machine learning methods have been proposed for risk classification. While machine learning methods have also been used for failure diagnosis and condition monitoring, to the best of our knowledge, these methods have not been used for probabilistic risk assessment. Probabilistic risk assessment is a subjective process. The problem of how well machine learning methods can emulate expert judgments is challenging. Expert judgments are based on mental shortcuts, heuristics, which are susceptible to biases. This paper presents a process for developing natural language-based probabilistic risk assessment models, applying deep learning algorithms to emulate experts’ quantified risk estimates. This allows the risk analyst to obtain an a priori risk assessment when there is limited information in the form of text and numeric data. Universal sentence embedding (USE) with gradient boosting regression (GBR) trees trained over limited structured data presented the most promising results. When we apply these models’ outputs to generate survival distributions for autonomous systems’ likelihood of loss with distance, we observe that for open water and ice shelf operating environments, the differences between the survival distributions generated by the machine learning algorithm and those generated by the experts are not statistically significant.     

Traffic Partition in WDM/SONET Rings to Minimize SONET ADMs

SONET (Synchronous Optical NETworks) add-drop multiplexers (ADMs) are the dominant cost factor in the WDM(Wavelength Division Multiplexing)/SONET rings. The number of SONET ADMs required by a set of traffic streams is determined by the routing and wavelength assignment of the traffic streams. Previous works took as input the traffic streams with routings given a priori and developed various heuristics for wavelength assignment to minimize the SONET ADM costs. However, little was known about the performance guarantees of these heuristics. This paper contributes mainly in two aspects. First, in addition to the traffic streams with pre-specified routing, this paper also studies minimizing the ADM requirement by traffic streams without given routings, a problem which is shown to be NP-hard. Several heuristics for integrated routing and wavelength assignment are proposed to minimize the SONET ADM costs. Second, the approximation ratios of those heuristics for wavelength assignment only and those heuristics for integrated routing and wavelength assignment are analyzed. The new Preprocessed Iterative Matching heuristic has the best approximation ratio: at most 3/2.     

Heuristics for single-row layout problems in flexible manufacturing systems

A series of algorithms relevant to one of the commonly used layout schemes in flexible manufacturing systems, that is, the single-row layout, is presented. The algorithms constitute a modified version of a known technique developed for the n m F Fmax sequencing problem. As a result, heuristics are obtained which are easy to implement and which generate good solutions in acceptable response times and outperform other algorithms developed to the same end. An iterative method based on a neighbourhood search is also developed to evaluate the performances in absolute terms.     

Heuristic Optimization of Physical Data Bases: Using a Generic and Abstract Design Model*

Designing efficient physical data bases is a complex activity, involving the consideration of a large number of factors. Mathematical programming-based optimization models for physical design make many simplifying assumptions; thus, their applicability is limited. In this article, we show that heuristic algorithms can be successfully used in the development of very good, physical data base designs. Two heuristic optimization algorithms are proposed in the context of a generic and abstract model for physical design. One algorithm is based on generic principles of heuristic optimization. The other is based on capturing and using problem-specific information in the heuristics. The goodness of the algorithms is demonstrated over a wide range of problems and factor values.     

Heuristics for scheduling in a flowshop with setup,processing and removal times separated

The problem of scheduling in a flowshop, where setup, processing and removal times are separable, is considered with the objective of minimizing makespan. Heuristic algorithms are developed by the introduction of simplifying assumptions into the scheduling problem under study. An improvement method is incorporated in the heuristics to enhance the quality of their solutions. The proposed heuristics and an existing heuristic are evaluated by a large number of randomly generated problems. The results of an extensive computational investigation for various values of parameters are presented.     

Distribution with Quality of Service Considerations: The Capacitated Routing Problem with Profits and Service Level Requirements

Inspired by a problem arising in cash logistics, we propose the Capacitated Routing Problem with Profits and Service Level Requirements (CRPPSLR). The CRPPSLR extends the class of Routing Problems with Profits by considering customers requesting deliveries to their (possibly multiple) service points. Moreover, each customer imposes a service level requirement specifying a minimum-acceptable bound on the fraction of its service points being delivered. A customer-specific financial penalty is incurred by the logistics service provider when this requirement is not met. The CRPPSLR consists in finding vehicle routes maximizing the difference between the collected revenues and the incurred transportation and penalty costs in such a way that vehicle capacity and route duration constraints are met. A fleet of homogeneous vehicles is available for serving the customers. We design a branch-and-cut algorithm and evaluate the usefulness of valid inequalities that have been effectively used for the capacitated vehicle routing problem and, more recently, for other routing problems with profits. A real-life case study taken from the cash supply chain in the Netherlands highlights the relevance of the problem under consideration. Computational results illustrate the performance of the proposed solution approach under different input parameter settings for the synthetic instances. For instances of real-life problems, we distinguish between coin and banknote distribution, as vehicle capacities only matter when considering the former. Finally, we report on the effectiveness of the valid inequalities in closing the optimality gap at the root node for both the synthetic and the real-life instances and conclude with a sensitivity analysis on the most significant input parameters of our model.     

Reducing the number of setups for CNC punch presses

Large computer numerical control (CNC) punch presses are widely used for punching holes in heavy gage sheet metals. The turret of such presses has limited tool capacity, and any tool changes require dismantling the turret. Dismantling and reloading the turret generally takes several hours and is the dominant element of the total flow time for a batch of sheet metals. Reducing the turret setup time requires judicious grouping of the required tool set. We formulated an integer program to obtain the minimum number of setups. However, since the problem is nondeterministic polynomial-time hard (NP-hard), we offer three heuristics to solve the problem. We compare solutions obtained from the three heuristics with their respective optimal solutions using three simulated data sets and an industrial data set.     

Ranking irregularities when evaluating alternatives by using some ELECTRE methods

The ELECTRE II and III methods enjoy a wide acceptance in solving multi-criteria decision-making (MCDM) problems. Research results in this paper reveal that there are some compelling reasons to doubt the correctness of the proposed rankings when the ELECTRE II and III methods are used. In a typical test we first used these methods to determine the best alternative for a given MCDM problem. Next, we randomly replaced a non-optimal alternative by a worse one and repeated the calculations without changing any of the other data. Our computational tests revealed that sometimes the ELECTRE II and III methods might change the indication of the best alternative. We treat such phenomena as rank reversals. Although such ranking irregularities are well known for the additive variants of the AHP method, it is the very first time that they are reported to occur when the ELECTRE methods are used. These two methods are also evaluated in terms of two other ranking tests and they failed them as well. Two real-life cases are described to demonstrate the occurrence of rank reversals with the ELECTRE II and III methods. Based on the three test criteria presented in this paper, some computational experiments on randomly generated decision problems were executed to test the performance of the ELECTRE II and III methods and an examination of some real-life case studies are also discussed. The results of these examinations show that the rates of the three types of ranking irregularities were rather significant in both the simulated decision problems and the real-life cases studied in this paper.     

An object-oriented methodology for solving the RCPSPs with heuristics and metaheuristics

Because various heuristics and metaheuristics have been proposed to solve the well known NP-hard, resourceconstrained project scheduling problem (RCPSP), it is currently difficult to compare the computational efficiency of these heuristics implemented on different computers where, in addition, the computer codes may have been written in different computer languages. This problem is solved when all relevant heuristics can be applied within the framework of a single computer program. By use of the object-oriented programming (OOP) methodology, we developed a general software framework for the heuristics and metaheuristics for solving the RCPSP. Currently this includes six heuristics and two metaheuristics. The framework of the software allows a more advanced user to append more effective heuristics and play around with several parameters of these metaheuristics with a bare minimum of coding effort.     

An improved MIP heuristic for the intermodal hub location problem

Optimization methods have been commonly developed for the intermodal hub location problem because it has a broad range of practical applications. These methods include exact methods (limited on solving large-size problems) and heuristics (no guarantee on solution quality). In order to avoid their weakness but to leverage their strength, we develop an improved MIP heuristic combining branch-and-bound, Lagrangian relaxation, and linear programming relaxation. In the heuristic, we generate a population of initial feasible solutions using the branch-and-bound and Lagrangian relaxation methods and create a linear-relaxed solution using the linear programming relaxation method. We combine these feasible and linear-relaxed solutions to fix a portion of hub location variables so as to create a number of restricted hub location subproblems. We then combine the branch-and-bound method to solve these restricted subproblems for iteratively improving solution quality. We discuss in detail the application of the method to the intermodal hub location problem. The discussion is followed by extensive statistical analysis and computational tests, where the analysis shows statistical significance of solutions for guiding the heuristic search and comparisons with other methods indicate that the proposed approach is computationally tractable and is able to obtain competitive results.     

A hybrid algorithm based on variable neighbourhood for the strip packing problem

This paper addresses the strip packing problem, which has a wide range of real-world applications. Our proposed algorithm is a hybrid metaheuristic that combines an improved heuristic algorithm with a variable neighbourhood search. Different neighbourhoods are constructed based on the concept of block patterns. The proposed algorithm has three interesting features. First, a least-waste strategy is used to improve the constructive heuristics. Second, a better sorting sequence is selected to generate an initial solution. Finally, different neighbourhoods are constructed based on block patterns. The computational results from a diverse set of problem instances show that the proposed algorithm performs better than algorithms reported in the literature for most of the problem sets compared.     

Overall material usage effectiveness (OME): a structured indicator to measure the effective material usage within manufacturing processes

A rational use of production resources has been catalysing the attention of researchers and practitioners for several years. A number of indicators have been proposed to assess the effective and efficient use of the various production factors. A widely known example of these metrics is the overall equipment effectiveness (OEE), which applies to machines and equipment. This paper proposes a novel methodology that seizes upon OEE’s straight forward and easy-to-use structure to address the problem of measuring the effective use of materials within a factory. An inclusive structure of material loss and a methodology to calculate a novel performance indicator called overall material usage effectiveness (OME) is given. The proposed set of KPIs not only permits the comprehension of material-related issues but also the identification of viable countermeasures. A real-life application to demonstrate the practical usefulness of the methodology and illustrate the capability of its KPIs is described.     

A memetic algorithm for the patient transportation problem

This paper addresses a real-life public patient transportation problem derived from the Hong Kong Hospital Authority (HKHA), which provides ambulance transportation services for disabled and elderly patients from one location to another. We model the problem as a multi-trip dial-a-ride problem (MTDARP), which requires designing several routes for each ambulance. A route is a sequence of locations, starting and terminating at the depot (hospital), according to which the ambulance picks up clients at the origins and delivers them to the destinations. A route is feasible only if it satisfies a series of side constraints, such as the pair and precedence constraints, capacity limit, ride time, route duration limit and time windows. Owing to the route duration limit, in particular, every ambulance is scheduled to operate several routes during the working period. To prevent the spread of disease, the interior of the ambulances needs to be disinfected at the depot between two consecutive trips. The primary aim of the problem investigated herein is to service more requests with the given resources, and to minimize the total travel cost for the same number of requests. In this paper, we provide a mathematical formulation for the problem and develop a memetic algorithm with a customized recombination operator. Moreover, the segment-based evaluation method is adapted to examine the moves quickly. The performance of the proposed algorithm is assessed using the real-world data from 2009 and compared with results obtained by solving the mathematical model. In addition, the proposed algorithm is adapted to solve the classic DARP instances, and found to perform well on medium-scale instances.