Seminars

Upcoming CSAM Events

• 25th February 2025. Dr Chaitra Nagaraja. Senior Lecturer Mathematics and Statistics, University of Exeter.
  • Title: Exploring the bullwhip effect in a small business setting
  • Abstract: To follow.
•  4th February 2025. Professor Nav Mustafee. Professor of Analytics and Operations Management.
  • Title: A Hybrid Modelling and Simulation for Supporting Decision-making for Industrial Symbioses
  • Abstract: To follow.
• 21st January 2025. Oli Young. PhD student.
  • Title: Exploring Consumer Experience through Psychophysiological Data: Insights into Informal Learning at Heritage Attractions
  • Abstract: To follow.
• 14th January 2025. Professor Andreas Wieland. Copenhagen Business School. More information to follow.
• 3rd December 2024. Dr Lina Zhang.  Lecturer in Operations Management at the University of Exeter Business School
  • Title: Blessing or Curse? Introduction of a Live-streaming Channel.
  • Abstract:  Problem definition: Manufacturers can deploy either their internal brand representatives (IN) or external influencers (EX) for live-streaming (LS) selling. These two LS channel strategies are widely observed in practice but not fully explored in the literature. We study consumer behaviour under these strategies to understand whether and how manufacturers should utilize an LS channel to gain benefits. Methodology: We develop a model to investigate each LS channel strategy compared to the benchmark without LS selling. We also compare and analyse these strategies to shed light on the conditions under which a manufacturer should employ IN or EX. Results: We uncover different mechanisms that drive the performance of each LS channel strategy. Compared to the benchmark, adopting IN benefits the manufacturer, given demand spill over effects, whereas EX may worsen the manufacturer. There are sufficient conditions to make the manufacturer better off in EX (than the benchmark), such as adequate cross-channel product-driven spill over effects and a reasonably high product valuation by influencer-driven consumers. Furthermore, we provide an influencer's aggregate authority threshold beyond which EX is preferable to IN, and we show how this threshold varies with problem parameters like the strength of product-driven spill over effects and influencer-driven consumers' product valuation. These findings are robust to endogenous or exogenous commission rates. Managerial implications: We provide important managerial insights for manufacturers to embrace LS e-commerce. Our results can help manufacturers make informed decisions regarding when and which LS channel strategy they should choose when working with different types of live streamers.
• 26th November, 2024. Professor Justin Tumlinson Associate Professor in Business Analytics, University of Exeter.
  • Title: Language as a Strategic Choice — Drawing Global Research Talent by Switching to English
  • Abstract: Organizations around the world are switching operations to English to ease global communication and tap global talent pools. Universities, aiming increase their research power, are amongst the most visible. However, concerns about domestic inequality, language requirements for local employers, and integration of immigrants have led to a feverish public debate surrounding academia’s embrace of the lingua franca in many countries like the Netherlands, Belgium, and China. This paper quantifies the benefits of introducing degree programs in English to university research performance and faculty recruiting. We show that, by introducing English as the official language of instruction in a degree program in the years 1991 to 2020, thirty Spanish, Dutch, Belgian, German, and Danish universities increased the academic ability of new faculty hires (as proxied by publication performance) by about 150% and their proportion of international new hires by about 24%.  These results are robust to a battery of controls, including university and time fixed effects, and university-specific time trends. We argue that switching to English reduces labor mobility barriers, allowing universities in non-English speaking countries to recruit from the global talent pool. Conversely, the long-standing cream-skimming opportunities of Anglo-American universities are threatened by the move. These findings add a quantitative dimension to the public debate of globalization in science.

Recent CSAM Events

• 5th November 2024. Dr Geng Cui. Project Research Associate, University of Tokyo.
  • Title: Reinforcement learning and inventory management
  • Abstract: Knowledge-based techniques, such as modeling and simulation, often require restrictive assumptions to ensure solvability and rely heavily on extensive domain expertise for both solving the models and deriving managerial insights. These limitations significantly hinder the ability of knowledge-based approaches to address practical tasks. However, the rapid development of artificial intelligence offers an alternative to overcoming these challenges. Reinforcement learning, a subfield of artificial intelligence, excels at developing optimal policies for sequential decision-making problems. Its key strength lies in not relying on human domain knowledge, but rather on computational resources to train a model. As a result, these algorithms are particularly suited for solving complex tasks that are difficult to address with human expertise and traditional methods alone. We assess the performance of reinforcement learning in inventory management, beginning with a simple linear model and extending to more complex scenarios, including a two-stage supply chain and inventory systems with temporary capacity constraints. Our findings show that reinforcement learning matches the performance of well-established knowledge-based techniques and, in some cases, may even outperform them.
• 22nd October 2024. Professor Nav Mustafee. Professor of Analytics and Operations Management, University of Exeter.
  • Title: Farmers' Perception of Drought Impacts and Strategies for Mitigation and Adaptation: A Case Study from India
  • Abstract: Climate change adversely affects impacts farming and agriculture in India. The study is from the Vidarbha region of Maharashtra, famous for its sweet-sour-flavoured "Nagpur Orange". The region has been witnessing droughts, resulting in crop failure and distress for growers of Nagpur oranges. The study attempts to identify critical challenges facing the orange growers, examine their perception and awareness about drought impacts, and different adaptation and mitigation strategies. A three-stage research design is implemented. In the first stage, a modified Delphi method is adopted to identify and validate orange growers' challenges due to drought impact. The second stage utilises multi-criteria decision-making (MCDM)-based best-worst method to prioritise and determine the most critical challenges. This is followed by a descriptive analysis of questionnaires (collected as part of a field study) to determine the overall perception of orange growers about drought and its adaptation and mitigation strategies. The work was inspired by the EU-funded ARSINOE project that aims to create climate-resilient regions through systemic solutions and innovations. A small funding proposal for collaboration with India was subsequently developed and funded through the University of Exeter Global Partnership research awards.
• 8th October 2024. Professor Stephen Disney. Professor of Operations Management, University of Exeter.
  • Title:  The dynamics of agricultural supply chains.
  • Abstract: Pork products account for one-third of all meat consumed; half of all pork originates from China. Traded as lean hogs, fattened, four-month-old piglets exhibit a highly variable sales price, often fluctuating +/- 40% within a few months. To pinpoint the source of this dynamic uncertainty, we build a stylized model of the pig supply chain using difference equations and z-transforms to study the dynamics of the pork production system.  The extended lead times, where output occurs in five batches between 16 and 40 months after the initial decision (to grow a breeding sow), lead to some counter-intuitive inventory behaviour that is contrary to normal inventory theory.
• 24th September 2024. Professor Leroy White. Professor of Management, University of Exeter Business School
  • Title:  A discussion on Research Culture.
• 10th September 2024. Dr Fanlin Meng, Senior Lecturer in Operations and Analytics, University of Exeter Business School
  • Title: Energy management and electricity market: a data analytics perspective
  • Abstract: In this talk, I will give an overview of my recent research in energy management and electricity market (e.g. demand response, wholesale and retail electricity market bidding and pricing, smart meter data analytics, etc.) over last five years as well as some ongoing work.  The methods I employ in this area generally range from modelling and optimisation (e.g. bilevel optimisation or Stackelberg game in game theory community, scenario-based stochastic programming), machine learning (e.g. unsupervised learning such as dynamic clustering, supervised learning such as privacy-aware distributed and online learning, federated learning and transfer learning, reinforcement learning), to the data-driven modelling and learning (combining modelling, optimisation and machine learning).

• 13th August 2024. Dr Qingying Li, Donghua University
  • Title: Competition between Service Platforms: Myopic vs. Farsighted Stability
  • Abstract: In platform economy, service providers often engage in both demand and supply competition. This competition can be either farsighted or myopic, depending on how the platforms anticipate the reactions of their rivals. Farsighted platforms consider the chains of reactions following their initial deviation and the long-term consequences of their actions, while myopic platforms only focus on the immediate payoff change of their one-step deviation. We study the farsighted stability in service competition between two service platforms and compare the stable outcome with the myopic one. We consider two scenarios: the platforms engage in demand competition only or in both demand and supply competition. For each scenario, we first determine the equilibrium of worker-customer subgame, and then consider the stable outcomes for the service pricing decisions. For the farsighted stability, we follow the concept of the von Neumann-Morgenstern farsightedly stable set. Specifically, we establish the Pareto efficiency and then verify the indirectly dominance to obtain the farsightedly stable outcomes. For the demand competition game, we find that the farsightedly stable outcome coincides with the myopic one in ample- and moderate-size markets but differ in the scarce-size market, and the farsighted outcomes can achieve the same efficiency as the centralized system. For the two-sided competition game, we find that farsighted platforms can coexist in the market with both earning a positive profit, while myopic platforms may lead to an equilibrium with zero profit for all. Farsighted platforms also tend to extract more surplus from both customers and workers than do myopic platforms. However, the platforms’ farsighted behaviour still falls short of achieving the optimal profit that a centralized system can attain.
• 19th July 2024. Yi Liu, PhD student University of Bath
  • Title: Trade time with shopping basket discounts: Evaluating customers’ willingness to accept slower deliveries
  • Abstract: Owing to intense competition in last-mile delivery, online retailers are compelled to compete in delivery speed by offering fast and ultrafast delivery options. However, this fast delivery has become a cost burden for retailers and price-sensitive customers. To address this issue, some retailers are exploring strategies that incentivize customers to opt for slower delivery, making it more cost-effective and sustainable. This innovative delivery option emphasizes providing discounts on product pricing, or shopping baskets, rather than mere compensation for delivery fee. However, to examine customers’ perceptions and retailers’ feasibility of this delivery option, we employ the Gabor–Granger model and three experiments to investigate how customers trade off longer delivery times with discount incentives on product pricing. Across all respondents, 76% of them could accept a discount of less than 15% for slow delivery service. Even though our findings indicate that customers expect larger discounts for longer delivery times, we show that, as delivery time increases, such discounts exhibit diminishing marginal utility, becoming proportionately smaller. Our findings also show that two of the key factors under examination – retail price and product category – have a significant impact on customers’ willingness to accept slower delivery. Further, our results elucidate that the default delivery times (i.e. the delivery option offered to customers for free) have no impact on customers’ perception towards slow delivery. Instead, customers would simply request discount compensation for the extra waiting time. We also find that compared with young women, older males are more willing to use the slower delivery. Based on these results, we provide insights for retailers as to how to incorporate slower delivery into their existing delivery options and the customers they should target for the slower delivery service.

• 2nd July 2024. Justin Tumlinson Associate Professor in Business Analytics, University of Exeter.
  • Title: Branded by Bankruptcy
  • Authors: Anna Grosman, Enrico Onali, Justin Tumlinson
  • Abstract: We examine the impact of corporate bankruptcies on the career trajectory of individuals without board positions. We provide evidence of a decrease in the probability of holding a top leadership position of around 2.1 percentage points (unconditional probability: 9.2%) and an increase in the waiting time before holding such positions by around seven years relative toother individuals. The probability of holding board positions within two years after bankruptcy, or to be promoted, is also lower than for other individuals, while junior positions are more likely. To enhance our identification strategy, we also present results exploiting the simultaneous bankruptcy of Lehman Brothers and the rescue of Merrill Lynch as a plausibly exogenous shock. Lehman’s bankruptcy leads to a lower probability of being promoted to a board position and a higher probability of being hired in a junior role, even for individuals who do not become jobless as a result of Lehman’s bankruptcy (e.g., Lehman’s US assets bought by Barclays or equities and investment banking operations in Europe, the Middle East, Africa and Asia bought by Nomura). These results are novel in that they show that corporate bankruptcies can slow the career advancement of individuals without board positions at the time of bankruptcy.

• 4th June 2024. Qing Zhu. Visiting PhD student.
  • Title: Optimal contract selection strategies under three different structures
  • Abstract: We explore optimal contract choice under monopoly, retailer competition, and manufacturer competition structures. The impact of interest rates and the substitution rate between two products on manufacturers' and retailers' profits is also considered. We also focus on the similarities and differences between the manufacturers and the retailers under Bertrand and Cournot competition. We identify the contracts the manufacturers wish to adopt, the contracts the retailers prefer manufacturers to choose, and whether there are differences in the reasons for contract selection. Our model employs a two-stage Stackelberg game. The research findings indicate that in the monopoly case,  with retailers engaging in either Bertrand and Cournot competition, the volatility of spot market price affects the manufacturer's contract choice and the retailers' contract preferences in the same manner. We found that under manufacturer competition, whether in Bertrand or Cournot games, the manufacturer's contract choices and equilibrium strategies are the same. However, in this structure, the  retailers' preferences for the type of contract chosen by manufacturers may vary depending on the magnitude of market demand price volatility. Furthermore, the study concludes that under the two game modes of manufacturer and retailer competition, both manufacturer and retailer profits decrease with an increase in the substitutability between the two products, and manufacturer profits decrease with interest rate increases.

• 9th April 2024. Yogendra Singh. PhD student.
  • Title: Reducing Bullwhip and MRP nervousness with propotional future guidance
  • Abstract: Material Requirements Planning (MRP) systems regulate raw material inventory (RMI) levels by periodically issuing call-off orders and order forecasts to suppliers. These orders and forecasts change with each regeneration of the MRP plan, often resulting in a costly effect called MRP nervousness. The manufacturer receives orders from customers, sets production targets, and issues replenishment orders to the supplier. A typical manufacturing echelon of a supply chain comprises of three lead times; the lead time experienced by the customer, the lead time required by the shop floor to produce finished goods inventory (FGI), and the lead time the supplier requires to deliver RMI. If a customer is willing to wait for the product, their advanced demand information (ADI) can lead to lower FGI levels/costs. However, how the ADI affects MRP nervousness is unknown. Using value stream maps and control theory, we model and analyse a manufacturing supply chain echelon to understand how the three lead times influence the MRP nervousness in the future order guidance given to the supplier.

• 26th March 2024.  Professor Dong Li. Professor of Operational Research, Lancaster University Management School.  
  • Title: Choice-based availability controls for urban carsharing revenue management
  • Abstract: Urban carsharing services provide flexible, affordable and green mobility options where customers can pick up a car from one station and return it to the same station (round trip) or any other station in the car-sharing network (one-way).  While one-way car-sharing offers a higher degree of mobility flexibility to customers, it may cause operational challenges such as fleet unbalance between multiple rental stations. Carsharing operators have employed various vehicle relocation strategies to ensure the availability of sufficient vehicles across rental stations. In this study, we adopt the user-based relocation strategy where customers are offered incentives (e.g., fare discounts) to drop off the vehicles to alternative destinations. We estimate a Multinomial logit (MNL) choice model based on collated responses from a discrete choice experiment implemented in a survey. We build the choice model into a dynamic program to determine the optimal destination-fare discount combinations offered to each arriving customer, which maximises the total expected revenue over a finite time horizon. To address the computational complexity of the dynamic program, we propose two approximation approaches: a choice-based deterministic linear program and a decomposition method. An extensive numerical study shows the effectiveness and efficiency of the proposed approaches, especially the latter.

• 20th March 2024.  Marta Staff. PhD student.
  • Title: Human Donor Milk-Potential Demand Estimation.
  • Abstract: Using donor human milk (DHM) for preterm infants, where the mother’s milk is unavailable, has multiple health benefits. Understanding the relationship between clinical choices for DHM provision and the resulting demand is important for multiple stakeholders. For policymakers, it informs decision-making around the provision of DHM based on cost-benefit analyses. For milk banks, it helps plan for required capacity, donor recruitment and supply-side collections. Beyond sharing the “latest and greatest” framework and model (and case study results) for demand estimation I will also share, with the audience, how the manuscript has evolved over time.

• 27th February 2024.  Dr Geng Cui. Project Research Associate, University of Tokyo.
  • Title: Jamology: Analysis and solution for various kind of jams
  • Abstract: Traffic jams are not only observed in vehicles but also in pedestrian flow and logistics. "Jamology" analyzes these phenomena using various approaches in a multidisciplinary manner. Eliminating traffic jams offers numerous benefits: it reduces traffic congestion and waiting times in vehicle transportation systems, lowers the risk of crowd accidents in areas with high pedestrian traffic, and decreases waste and improves production efficiency in logistics systems. In this presentation, I will introduce several research topics covered within the scope of Jamology, including vehicle traffic, pedestrian dynamics, and logistics. The approaches utilized in Jamology studies encompass statistical processes, fluid dynamics, and artificial intelligence. I will begin by describing the application of cellular automata to vehicular transportation research, then transition to pedestrian dynamics related research, and finally describe my study in the field of logistics.

• 13th February 2024. Dr Alison Harper. Lecturer in Operations & Analytics, Exeter Business School.

  • Title: Open-source Modeling for Orthopaedic Elective Capacity Planning using Discrete-event Simulation

  • Abstract:  Increases in elective surgical waiting lists over the last few years are creating significant consequences for health services.  The allocation of NHS capital funds to increase capacity for managing elective waits has created planning and operational challenges for health services. This presentation describes the development and deployment of an interactive web-based discrete-event simulation model for supporting capacity planning of surgical activity and ward stay in a proposed new ring-fenced orthopaedic facility in a UK health service. The model is free and open-source and developed to be generic and applicable for new capacity planning of elective recovery in orthopedics in other regions.

• 30th January 2024.  Dr Xuan Vinh Doan. Reader at Warwick Business School.
  • Title: Operations research games under uncertainty: a robust chance constrained approach
  • Abstract: Operations research games such as linear production games concern the cooperation among players who would face a joint optimization problem and share benefits/costs together if they agree to cooperate. In this talk, we aim to address a fundamental challenge of incorporating uncertainty into these games. We introduce a new solution concept of (robust) least chance decisions under uncertainty and distributional ambiguity, which is motivated by the concept of least core solutions for deterministic games. We develop a framework to find those decisions and compute their (robust) least chance dissatisfaction for games under normally distributed uncertainty and moment-based distributional ambiguity. We demonstrate how the framework can be applied to operations research games such as linear production games with detailed analytical results.

• 5th December 2023.  Dr Avril Sun. Lecturer in Operations and Supply Chain Management, University of Exeter.
  

  • Title: Supply Chain Resilience: The Case of Chinese Pig Industry. 

  • Abstract. The focus on lean production in turbulent business environment and increasing unforeseen adverse events have resulted a higher likelihood of severe supply chain disruptions, which drives the need for supply chain resilience (SCR) building as a mean for organizations to achieve business continuity through adapting, responding, and recovering from unexpected risks. Responding to the call to better understand the mechanism of SCR building and framed by the contextual issues arising from the agricultural supply chain, the overarching purpose of this study is to advance knowledge in supply chain resilience beyond the traditional risk management and static resilience approach. More specifically, this study explores how organizations make decisions on resources investment and build SCR capabilities during risk recovery and discovered how supply chain co-evolves with its environment in terms of SCR building through analysing organization’s resources investment configuration strategies. This study presents empirical findings drawn from multiple case studies within four Chinese leading pig production organisations, focusing on their risk mitigation processes. Building on conservation of resources theory and ambidexterity theory, a novel resilience building framework was revealed, and a theory of Resilient Resource Based View (RRBV) was developed.

• 7th November 2023. Dr Justyna Rybicka, Technology Specialist at the Manufacturing Technology Centre.
  • Title: From mathematical model to digital twin in manufacturing– best practice for modelling and simulation.
  • Abstract: This session aims, through lens of best practice in development of the operational simulation models, showcase use cases from the cross-sector manufacturing applications. It will cover the modelling simulation capabilities, model development best practice and practical use cases showcasing different simulation examples for application at different level of technology readiness levels.

• 25th October 2023. Professor Nenad Jukic, Quinlan School of Business, Loyola University Chicago. 
  • Title: Prevalence of Conceptual Modeling in Database Development Projects
  • Abstract: Conceptual database modeling is widely accepted in academia as an essential and integral part of the database development process, essential for minimizing the risk of information systems project failures.  It is described in detail in database textbooks, and it is taught at most university level database courses at the undergraduate and graduate levels.  However, there has not yet been a research study published that examines the prevalence of conceptual database modeling in real world database projects.  The ongoing research that will be described in this talk aims to conduct such a study.  In the contemporary world, daily interaction with databases is a fact of life. It occurs in many everyday situations, such as when we make purchases, view content on the internet, or engage in banking transactions. The purpose of a database is to organize the data in a way that facilitates straightforward access to the information captured in the data. The first step in the development of databases is requirements collection, definition, and visualization.  This step results in the end user requirements specifying which data the future database system will hold, and in what fashion, and what the capabilities and functionalities of the database system will be. The requirements are used to model and implement the database. If this step is successful, the remaining steps have a great chance of success. However, if this step is done incorrectly, all the remaining steps, and consequently the entire project will fail.  The database requirements step has three parts: requirement collection, requirement definition, and requirements visualization (also known as conceptual modeling).  Requirement collection refers to the process of eliciting requirements from the clients and/or future users of the database (typically via conversations and interviews).  Requirement definition refers to recording the collected requirements into a written document.  Requirement visualization/conceptual modeling refers to creating a blueprint for the database which is a direct reflection of the requirements.  To illustrate conceptual modeling, in this talk a brief demonstration will be given using a small example of developing a database. The example will be conducted using ERDPlus, a free web-based database modeling suite created by Prof. Jukic, used daily by thousands of academic and industry users worldwide. Following the demonstration, the ongoing survey of database professionals will be presented.  The survey identifies the levels of use of conceptual database modeling, and it investigates the reasons for not using conceptual database modeling among the practitioners who do not typically use it. The survey also examines how usage (or lack thereof) of conceptual modeling affects the overall outcome of the database development process. The initial results of this survey, based on the responses of several hundred database practitioners, will be shared.   
• 11th October 2023. Professor Justin Tumlinson. Associate Professor in Business Analytics, University of Exeter.
• 27th September 2023. Professor Nav Mustafee. Professor of Analytics and Operations Management, University of Exeter.
  • Title: Hybrid models with real-time data: Characterising real-time simulation and digital twins
  • Abstract: Real-time Simulation (RtS) and Digital Twins (DT) are terms generally associated with hybrid models that use real-time data to drive computational models. Additionally, in the case of DTs, real-time data is often used to create virtual replicas of the physical system as it progresses through real-time. There is an increasing volume of literature on RtS and DT; however, the field of OR/MS is yet to coalesce on accepted definitions and conceptualisations. This has arguably led to the cascading usage of these terms. The objective of the paper is threefold: (1) distinguish between RtS and DT, (2) present RtS-DT conceptualisation in four dimensions, and (3) present methodological and technical insights on developing RtS with limited data. We argue that the evolution of conventional simulation models to fully-fledged hybrid DTs may necessitate a focus on a transitional stage; namely, RtS models primarily driven using historical distributions with limited real-time data feeds.