Operations Research Tools Research Articles

Structured pruning of neural networks for constraints learning

In recent years, the integration of Machine Learning (ML) models with Operation Research (OR) tools has gained popularity in applications such as cancer treatment, algorithmic configuration, and chemical process optimization. This integration often uses Mixed Integer Programming (MIP) formulations to represent the chosen ML model, that is often an Artificial Neural Networks (ANNs) due to their widespread use. However, ANNs frequently contain a large number of parameters, resulting in MIP formulations impractical to solve. In this paper we showcase the effectiveness of a ANN pruning, when applied to models prior to their integration into MIPs. We discuss why pruning is more suitable in this context than other ML compression techniques, and we highlight the potential of appropriate pruning strategies via experiments on MIPs used to construct adversarial examples to ANNs. Our results demonstrate that pruning offers remarkable reductions in solution times without hindering the quality of the final decision, enabling the resolution of previously unsolvable instances.

Models for optimising the volume of material flows in the technological chain of corporate vertically integrated structures of the agricultural sector

The relevance of this study lies in the need to optimise supply cycles and volumes in value chains, which helps to reduce costs and increase the profitability of agricultural enterprises. The purpose of this study was to investigate and optimise the costs of initial material flows in the production subsystems of corporate vertically integrated structures of the agricultural complex under conditions of non-stationary demand. To fulfil this purpose, the study investigated the relationship between the amount of raw material stock stored in the production subsystem and the time of its consumption, based on which an extended model of the economic order quantity (EOQ) was considered, which, apart from the defined costs, also considers the costs of raw material shortages associated with the inability to fulfil deliveries and losses associated with supply failures caused by the unpredictability (stochasticity) of the order flow itself. It was found that in continuous production systems there is an opportunity to reduce the costs associated with failures by using an additional regular supply batch. For this, the mathematical “point-of-order” model was built based on the theory of mass service, which allows determining not only the best point to order, but also the optimal amount of the safety stock. A model for optimising the volume of material flows was proposed, which combines the model of the economic order quantity order adapted for use in product subsystems of corporate integrated structures of the agricultural complex and the “point-of-order” model, which allows calculating the minimum size of the insurance stock of raw materials using the tools of operations research. As an example, the material flow was optimised for Kivshovata Agro LLC. The findings of the study, such as the use of analytical tools and models for determining the economic order quantity and safety stock, can be used by the management of agricultural enterprises to improve the efficiency of material flow management

Robust optimization of a radiotherapy pretreatment preparation workflow

Objective. Increasing cancer incidence, staff shortage and high burnout rate among radiation oncologists, medical physicists and radiation technicians are putting many departments under strain. Operations research (OR) tools could optimize radiotherapy processes, however, clinical implementation of OR-tools in radiotherapy is scarce since most investigated optimization methods lack robustness against patient-to-patient variation in duration of tasks. By combining OR-tools, a method was developed that optimized deployment of radiotherapy resources by generating robust pretreatment preparation schedules that balance the expected average patient preparation time (F mean) with the risk of working overtime (RoO). The method was evaluated for various settings of an one-stop shop (OSS) outpatient clinic for palliative radiotherapy. Approach. The OSS at our institute sees, scans and treats 3–5 patients within one day. The OSS pretreatment preparation workflow consists of a fixed sequence of tasks, which was manually optimized for radiation oncologist and CT availability. To find more optimal sequences, with shorter F mean and lower RoO, a genetic algorithm was developed which regards these sequences as DNA-strands. The genetic algorithm applied natural selection principles to produce new sequences. A decoder translated sequences to schedules to find the conflicting fitness parameters F mean and RoO. For every generation, fitness of sequences was determined by the distance to the estimated Pareto front of F mean and RoO. Experiments were run in various OSS-settings. Main results. According to our approach, the expected F mean of the current clinical schedule could be reduced with 37%, without increasing RoO. Additional experiments provided insights in trade-offs between F mean, RoO, working shift length, number of patients treated on a single day and staff composition. Significance. Our approach demonstrated that OR-tools could optimize radiotherapy resources by robust pretreatment workflow scheduling. The results strongly support further exploration of scheduling optimization for treatment preparation also outside a one-stop shop or radiotherapy setting.

A cascading model of stakeholder engagement for large-scale regional development using structured dialogical design

We present a large-scale systemic intervention conducted in the wine village communities of Cyprus. The scientific aim of this paper is to contribute to expanding community operational research's purpose, scope, and tools by reflecting on a cascading engagement model of stakeholders toward regional development in Cyprus. An ad-hoc committee comprising three representatives of the Department of Town Planning and Housing, three members of the Cyprus Academy of Public Administration, and a few external consultants came up initially with about a dozen stakeholders’ categories they considered “relevant.” Using a stepwise (cascading) expansion, they eventually engaged over 200 actors representing 26 stakeholder categories. The first 29 stakeholders identified 71 obstacles to developing the wine villages using Structured Democratic Dialogue (SDD). They clustered the obstacles into themes and formed working groups to address them. Each group identified and invited experts and additional stakeholders they considered relevant to form extended theme-specific groups. Each group conducted a SWOT analysis to deepen understanding and produced a thematic vision map using SDD. A holistic vision map was constructed through a final SDD process in which participants were the authors of the most influential factors from each thematic vision map. The approach adheres to the bottom-up, community-led method and falls within community operational research and citizen science. The cascading model offers a formula for multi-methodological intervention. The stepwise bottom-up identification and engagement of stakeholders empowers and engages authentically, especially those marginalised. The expansion in numbers and the effective engagement of stakeholders creates the required momentum for societal change. We discuss the model's broader applicability, frame it within third-phase science, and suggest future directions.

Foresite of Restaurant Business Development

Topicality. Domestic economic environment has been in a state of turbulence since the start of the coronavirus pandemic, additionally reinforced by the risks of a full-scale war escalation. The uncertainty of the crisis timeframe, as well as significant threats of physical destruction and financial losses cause the necessity of searching reference points for adaptation and revitalization of activities. The sphere of service activity, in particular the restaurant business, turned out to be the most vulnerable to the external influence of political and legal, social and cultural, technological tendencies, as well as fluctuations in economic conditions. The significance of this study is argued by the latter’s contribution to overcoming the shock of the Russian military invasion, which became decisive and drew attention to its social and economic role not only as an organizer of catering and leisure for the population, but also as a guarantee of food security. Therefore, the development of the restaurant business is an important aspect of supporting the country’s economy both in war and post-war realities. The prospects of startups should be supported by thorough calculation regarding the high level of its investment attractiveness. Therefore, the relevance of this research lies in the elaboration of oriented principles and prospects for the restaurant business development, which is based on scientific assessments, in particular thanks to foresight. The aim of the article. The aim of this study is to conduct a foresight research of the restaurant business in Ukraine, and to substantiate the oriented principles for development, taking into account market trends and tendencies. The specified topic determined the use of general scientific methods, which contributed to the elaborating key definitions and forming operational research tools. Research methods. The horizon scanning method was used to evaluate the restaurant business development scenarios; it revealed the main tendencies and trends of the restaurant services market and deterministic consumers’ needs. The modelling method helped to determine how communications could be formed in the process of implementation of foresight sessions for the restaurant business development. To assess the prospects for the development of the restaurant business, aspects of the pre-crisis (pre-quarantine and pre-war) as well as expectations (customers’, investors’ and subjects’ of the restaurant business) of the post-war periods were compared. Results. The key term “foresight” and its main predictors were defined. The components of the complex monitoring and prognostication mechanism were worked out: political, economic, technological and market foresight. An overview of the most important tendencies and trends in the national and global business environment was presented in order to identify and detail the problems and challenges facing restaurant establishments in future. Prospects for further research base on the elaboration of alternative scenarios for the restaurant business development through the foresight methodology. Conclusions and discussion. Restaurant business foresight studies use sophisticated methods, still, they rarely involve all stakeholders, including consumers. The authors pay special attention to the monitoring of political, economic, technological and market environment, as well as to digital technologies, which are already significantly changing approaches to business development. The analysis of Ukrainian subjects of the restaurant business makes it possible to evaluate strategic maps as a foresight tool, which is complemented by active prognostic methods. Based on the peculiarities of the restaurant business and market tendencies, four scenarios are defined: crisis experience (“freezing and adaptation”); anti-crisis benchmarking (“global expectations”); breakthrough innovations (“motivational attack”); resuscitation scenarios (“struggle and recovery”). All these scenarios describe oriented principles for the development of the restaurant business. They are practical for scientists and practitioners, as they provide a basis for elaborating and evaluating development strategies for both individual business entities and the restaurant industry in total.

Electric Vehicles and the Use of Demand Projection Models: A Systematic Mapping of Studies

In today’s world, electric vehicles have become a real solution to the problem of pollution caused by petrol and diesel-powered vehicles. However, incorporating them successfully into the global vehicle park poses new challenges. Some of these challenges have to do with meeting the electricity demand, providing the physical installations for charging, and the size and capacity of the electric grid required to deliver the necessary supply. Solving these new problems requires determining or projecting the electrical and/or physical requirements involved, but there is no single model or methodology to do this, nor any single document which summarizes the existing information. To address this situation, this work presents the result of a systematic mapping study that seeks to provide organized information about the (mathematical) models for the demand arising from electric vehicles, as well as to answer a series of questions posed for this research. The results obtained show that there is a wide variety of models used to determine demand requirements –of either physical or electrical elements– in which mathematical modelling and operations research tools are normally used. Other results indicate that demand models are mainly focused on the electrical requirements rather than on physical ones, and that, in most cases, the type of vehicle for which the demand is studied is not mentioned.

Application of an optimization software to a real-life case: Optimization of the partial electrification of a railway node

Within the transport accounting for 30% of GHG emissions in France, more than 2500 freight, shunting or passenger vehicles still run-on diesel in France. To reduce the carbon footprint of rail transport, battery-powered trains coupled with partial electrification of some lines could be a less expensive alternative to full electrification for low-traffic lines. In this paper, we present the interest of using an optimization software as an operational tool for a real use case of partial electrification of several lines in a territory. The systemic approach and the technical-economic approach of this operational research tool leads to a solution that minimizes the overall costs of decarbonization of a railway line from the infrastructure and rolling stock point of view: the electrification of 11% of non-electrified lines coupled with the battery train.

The use of modeling and simulation in the development process optimization of the Armed Forces of the Czech Republic

The article contributes to a more efficient use of the theoretical, technological and application base, especially to the improvement of decision-making processes at the strategic level. It offers a methodological framework for solving the armed forces capability optimization problem by implementing modeling, simulation and operations research tools and theories.

Transportation and Location Planning During Epidemics/Pandemics: Emerging Problems and Solution Approaches

The sudden changes in human mobility, the immense increase in demand for logistics and delivery systems, governmental restrictions, and uncertainty of the spread dynamics have introduced several transportation and location-related decision problems during the COVID-19 pandemic. Hence, a variety of Operations Research (OR) tools and techniques have been applied to tackle these problems for mitigating the adverse effects of the spread. In this study, we first investigate the emerging decision problematics observed during epidemics/pandemics under four research clusters as: ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${i}$ </tex-math></inline-formula> ) effects of epidemics on transportation, (ii) effect of mobility on pandemic spread, (iii) logistics and delivery systems, and (iv) medical waste management and wastewater-based epidemiology. Next, we explore the OR tools implemented to solve the transportation and location-related decision problems in each cluster.

Modelling Distribution Routes in City Logistics by Applying Operations Research Methods

The article focuses on the up-to-date subject from the practical as well as scientific point of view. It specifically discusses a proposal of an approach concerning transport or distribution problems in the range of city logistics and investigates possibilities to use opted operations research methods in this particular area. Specific suggestions lie first and foremost in using selected tools of operations research (i.e. a set of methods concerning vehicle routing problem) to model multiple variants of distribution paths from a determined hub to multiple spokes in order to minimise the overall travelled distance in an urban area. As far as the very research goes, to define distribution paths to supply multiple logistics objects in the range of city logistics, ensuing methods are step by step used: Clarke-Wright algorithm, Mayer algorithm and the nearest neighbour algorithm. The article consists of a conceptual section, describing the relevant theory as well as data and methods used, the practical part and the section encompassing an assessment of the key findings, along with the discussion. A suitable combination of adequate operations research methods and their application to city logistics issues is where an innovative solution of this research lies.

Business Model Innovation to Address Vegetable Supply Chain Issues: A Case Study of an Indian Startup

Purpose: Agricultural supply chains witness several challenges, which result in increased cost of operations, high wastages, and low efficiency. In the case of vegetable supply chains (VSCs), the intensity of challenges is very high due to perishability. Literature suggests various operations research tools for supply chain management of perishables, but there is a paucity of research highlighting the real models adopted by real players in the market while making revenue. This paper investigates and presents how a startup can innovate its business model to meet the VSC issues by adopting a single case study methodology. Design: The phenomenon was studied by convergent interviewing with the company’s key personnel and gathering information from the company’s key documents followed by content analysis. Findings: The study shows real practices adopted by a startup to cater to the challenges of VSCs in India by incorporating various tools of productivity enhancement and waste reduction. It highlights the adoption of a traceability system, production planning mechanism, reduced food miles, and green initiatives contributing to an innovative business model. The study maps the business model of the startup with the business model concept provided by Lindgardt and colleagues [ 2009 ] and states two more dimensions, i.e. “Client relationship management” and “Sustainability activities” to expand the mentioned configuration. Originality: The study is the first of its kind to highlight the Business Model Innovation of a startup to address the VSC challenges in the Indian context. Practical contribution: The study has important managerial implications useful for finding our new value propositions in a VSC firm by shortening and effectively integrating the supply chain while minimizing information asymmetry to achieve economies and sustainability. Theoretical contribution: This study has strengthened the methodological advances in the study of business models based upon earlier validated configurations. It contributes to the extant research by adding two more elements “Client relationship management” and “Sustainability activities” in a business model configuration.

Data analytics during pandemics: a transportation and location planning perspective.

The recent COVID-19 pandemic once again showed the value of harnessing reliable and timely data in fighting the disease. Obtained from multiple sources via different collection streams, an immense amount of data is processed to understand and predict the future state of the disease. Apart from predicting the spatio–temporal dynamics, it is used to foresee the changes in human mobility patterns and travel behaviors and understand the mobility and spread speed relationship. During this period, data-driven analytic approaches and Operations Research tools are widely used by scholars to prescribe emerging transportation and location planning problems to guide policy-makers in making effective decisions. In this study, we provide a review of studies which tackle transportation and location problems during the COVID-19 pandemic with a focus on data analytics. We discuss the major data collecting streams utilized during the pandemic era, highlight the importance of rapid and reliable data sharing, and give an overview of the challenges and limitations on the use of data.

Artificial intelligence and end user tools to develop a nurse duty roster scheduling system

ObjectivesA nurse duty roster is usually prepared monthly in a hospital ward. It is common for nurses to make duty shift requests prior to scheduling. A ward manager normally spends more than a working day to manually prepare and subsequently to optimally adjust the schedule upon staff requests and hospital policies. This study aimed to develop an automatic nurse roster scheduling system with the use of open-source operational research tools by taking into account the hospital standards and the constraints from nurses. MethodsArtificial intelligence and end user tools operational research tools were used to develop the code for the nurse duty roster scheduling system. To compare with previous research on various heuristics in employee scheduling, the current system was developed on open architecture and adopted with real shift duty requirements in a hospital ward. ResultsThe schedule can be generated within 1 min under both hard and soft constraint optimization. All hard constraints are fulfilled and most nurse soft constraints could be met. Compared with those schedules prepared manually, the computer-generated schedules were more optimally adjusted as real time interaction among nurses and management personnel. The generated schedules were flexible to cope with daily and hourly duty changes by redeploying ward staff in order to maintain safe staffing levels. ConclusionsAn economical but yet highly efficient and user friendly solution to nurse roster scheduling system has been developed and adopted using open-source operational research methodology. The open-source platform is found to perform satisfactorily in scheduling application. The system can be implemented to different wards in hospitals and be regularly updated with new hospital polices and nurse manpower by hospital information personnel with training in artificial intelligence.

Construction Operations Practices and Operations Research Tool Application in a Hypothetical Construction Project

This study aimed to assess the construction operations practices of three contractors; apply the construction operations practices for each company to a hypothetical construction project; and use an operation research tool in the hypothetical construction project. The results showed that the majority of contractors have similar organizational delivery systems with functional departments such as business development, finance, human resources, and construction operations, and they fall into the PCAB group. According to respondents, efficiency in resource use, avoidance of non-conformance works, elimination of wastes or non-value-added activities, value engineering, bulk order discounts, listing of different suppliers, double shifting of work scheme, prioritizing activities suitable for weather conditions, delegating work packages to subcontractors, resequencing the project schedule, application of value engineering are all important factors in optimizing project costs. The study concludes that operations research methods are important in project management, particularly in terms of reducing construction costs and increasing project profit. The findings indicated how the operations research tool, in conjunction with the QM for Windows application, made determining project schedule priorities to manage large and complicated project operations appear simple.

Optimizing Routes and Schedules of Occupational Health and Safety Specialists

According to occupational health and safety law of 6331, companies employing the certain number of workers or focusing the certain level of hazardous works must employ a full-time occupational health and safety (OHS) professional or be serviced by a part time one for prespecified durations. Many companies prefer to be serviced by part-time specialists since it is more cost-effective. Specialists working for independent joint health and safety units visit companies and service during predefined times daily basis by starting and ending the tours from/to their homes or offices. They spend long times among visits in traffic. Therefore, minimizing travel distances of specialists creates leisure times for rest and service. This problem is very similar to the Vehicle Routing Problem that has already been studied and applied many areas varied from logistic to security for long years. In this study, daily routing and scheduling activities of a specialist based on three scenarios are modelled with the mixed integer linear programming, frequently used to solve VRPs. Optimal routes and schedules for three scenarios are demonstrated and discussed. The main aim of the study is to show decision makers and specialists in the OHS area how to use operations research tools in their daily basis activities.

Computationally Efficient Approximations for Distributionally Robust Optimization Under Moment and Wasserstein Ambiguity

Distributionally robust optimization (DRO) is a modeling framework in decision making under uncertainty in which the probability distribution of a random parameter is unknown although its partial information (e.g., statistical properties) is available. In this framework, the unknown probability distribution is assumed to lie in an ambiguity set consisting of all distributions that are compatible with the available partial information. Although DRO bridges the gap between stochastic programming and robust optimization, one of its limitations is that its models for large-scale problems can be significantly difficult to solve, especially when the uncertainty is of high dimension. In this paper, we propose computationally efficient inner and outer approximations for DRO problems under a piecewise linear objective function and with a moment-based ambiguity set and a combined ambiguity set including Wasserstein distance and moment information. In these approximations, we split a random vector into smaller pieces, leading to smaller matrix constraints. In addition, we use principal component analysis to shrink uncertainty space dimensionality. We quantify the quality of the developed approximations by deriving theoretical bounds on their optimality gap. We display the practical applicability of the proposed approximations in a production–transportation problem and a multiproduct newsvendor problem. The results demonstrate that these approximations dramatically reduce the computational time while maintaining high solution quality. The approximations also help construct an interval that is tight for most cases and includes the (unknown) optimal value for a large-scale DRO problem, which usually cannot be solved to optimality (or even feasibility in most cases). Summary of Contribution: This paper studies an important type of optimization problem, that is, distributionally robust optimization problems, by developing computationally efficient inner and outer approximations via operations research tools. Specifically, we consider several variants of such problems that are practically important and that admit tractable yet large-scale reformulation. We accordingly utilize random vector partition and principal component analysis to derive efficient approximations with smaller sizes, which, more importantly, provide a theoretical performance guarantee with respect to low optimality gaps. We verify the significant efficiency (i.e., reducing computational time while maintaining high solution quality) of our proposed approximations in solving both production–transportation and multiproduct newsvendor problems via extensive computing experiments.

GRASP‐ILS and set cover hybrid heuristic for the synchronized team orienteering problem with time windows

AbstractWildfires are a natural phenomenon that regularly occurs in many terrestrial ecosystems. Due to global warming, the rate and the span of wildfires have remarkably increased during the last years, causing important economic losses and human casualties. Several initiatives have been undertaken in the last years in order to apply operations research tools to help firefighting teams schedule and optimize their protection activities when dealing with wildfires. In this context, a recent variant of the Team Orienteering Problem, referred to as the Asset Protection Problem, was proposed in van der Merwe et al. (2015). In this problem, firefighting teams provide a protective service to a set of assets endangered by wildfires. These activities can be performed by a heterogeneous fleet of vehicles and occur within specific time intervals estimated on the basis of fire fronts progression. This problem incorporates three additional constraints: time windows, synchronized visits, and compatibility constraints between vehicles and assets. In this paper, we propose a hybrid approach that combines a Greedy Randomized Adaptive Search Procedure coupled with an Iterated Local Search (GRASPILS) and a post‐optimization phase based on a set covering formulation. Interestingly, GRASPILS incorporates an adaptive candidate list‐based insertion heuristic and a Variable Neighborhood Descent search procedure. Detailed computational tests were carried out on benchmark instances from the literature. The results show that our method outperforms the other methods in the literature, since it improves all the best‐known solutions on medium‐ and large‐size instances, while maintaining shorter computational times.

A study on Efficient Modelling in Higher Education Academic Workforce Using Simulation

This paper shows how the Operations Research (OR) tools (in modelling and simulation) can be modified, applied in planning and their understanding of any long-term impacts due to sudden policy changes. The proposed approach is particularly useful to investigate the movements of the university academics and their impacts on changes in research, research funding, teaching and services as they are the integral parts of the career at any level. We argue that the Discrete Event Simulation (DES) approach can be used to model such dynamics in Higher Education Academic Workforce Model (HEAWM) and show that it can provide a comprehensive projection of future requirements within the context of career progression. Consequently, this HEAWM allows universities to interrogate factors influencing the academic workforce planning as this process often requires the new attributes to be tracked in the model which is difficult with other OR models. It is shown that this approach is easy to apply via DES and creation of the corresponding HEAWM provides better understanding of the factors that will influence the future workforce than the existing results.

Feature Selection for Recommender Systems with Quantum Computing.

The promise of quantum computing to open new unexplored possibilities in several scientific fields has been long discussed, but until recently the lack of a functional quantum computer has confined this discussion mostly to theoretical algorithmic papers. It was only in the last few years that small but functional quantum computers have become available to the broader research community. One paradigm in particular, quantum annealing, can be used to sample optimal solutions for a number of NP-hard optimization problems represented with classical operations research tools, providing an easy access to the potential of this emerging technology. One of the tasks that most naturally fits in this mathematical formulation is feature selection. In this paper, we investigate how to design a hybrid feature selection algorithm for recommender systems that leverages the domain knowledge and behavior hidden in the user interactions data. We represent the feature selection as an optimization problem and solve it on a real quantum computer, provided by D-Wave. The results indicate that the proposed approach is effective in selecting a limited set of important features and that quantum computers are becoming powerful enough to enter the wider realm of applied science.

Understanding housing preferences of slum dwellers in India: A community-based operations research approach

A significant number of India's urban population living in slums face large-scale challenges to access adequate housing and essential services. However, the solutions implemented by government agencies have been inadequate as they fail to understand the diversity of challenges and preferences of slum residents. While the traditional prescriptive approaches to study slum communities do not capture the uncertainties that riddle slums, we explore this domain of inquiry alternatively through theories of Community Operational Research (COR) and Community Based Operations Research (CBOR) that develop an understanding of housing problems from slum resident perspectives. In this paper, we study slum communities’ housing priorities by learning how they structure their concerns and identify specific solutions that could enable access to improved housing facilities. This is one of the first studies to apply Strategic Options Development & Analysis (SODA) and Value Focused Thinking (VFT) to engage with slum residents and generate in-depth insights into the lives of slum communities. Our findings from a slum in the Indian state of Odisha highlight the diverse challenges and needs of the community with regard to housing and basic infrastructure facilities from the slum-dwellers’ perspective, a voice often missing in slum policymaking. We believe that our findings could inform policymakers about most valued preferences of slum residents among many possible slum upgrading interventions. The study contributes to the extension of Operations Research tools and methodologies for meaningful engagement of vulnerable communities to develop interventions for improving social welfare.