Academic-practitioner bazaar

TA-56, Tuesday, 5th July, 8:30-10:00 - Building CW, 1st floor, Room 122 / Lobby

Please join us for an exciting line‐up of ac‐prac projects and ideas.

This is a unique opportunity:
- for practitioners to find an academic collaborator to help with a problem;
- for academics to find an industrial partner to apply theoretical developments.

The bazaar is a single 90‐minute session, during which each poster‐presenter gets 60 seconds to promote their poster to the whole audience, followed by an opportunity for people to visit the posters they are interested in, at a more leisurely pace, and have a discussion with the author(s).

If you have an idea for a collaborative project, but do not have time to prepare a poster, there is still an opportunity to provide a 60 second pitch during the session. To do this please sign up at the Making an Impact (MAI) stand in the CW exhibition area by Monday 4th 17.45.

Posters currently planned include:

1. Autonomous Dial‐a‐Ride Transit Systems: Optimization and adaptive network design strategies for large complex networks, Bongiovanni Claudia, Mor Kaspi, Nikolas Geroliminis
In this work we present new strategies for the design and management of Demand Responsive Transit Systems (DRTs) for autonomous vehicles. The design of autonomous vehicles to be integrated with DRTs is currently under ongoing work for various Swiss cities. In OR literature, a class of models have been developed to deal with such systems, all of which are generalizations of the well‐known Traveling Salesman Problem. In the dynamic Autonomous Dial‐A‐Ride Problem with Time Windows (ADARPTW) the goal is to optimize the dynamic routing and scheduling of a fleet of capacitated autonomous vehicles in an urban network in which demand changes over time and space with no a priori knowledge. The ADARPTW has also been extensively investigated for the static case, in which demand is completely known in advance. In both cases, the optimization can be performed considering operational and quality costs, finding the trade‐off between the agency's and users' interests. Since the ADARPTW is NP‐hard, several strategies have been proposed in order to deal with large instances, including the cluster‐first‐route‐second strategy. In this work real data from Uber Technologies Inc. in San Francisco is employed in order to test machine learning strategies to find best practices in clustering for the DARPTW in large complex networks.

Claudia Bongiovanni is a current Ph.D. candidate at the LUTS laboratory at EPFL. She has a B.Sc. degree in Civil and Environmental Engineering from the University of Roma Tre and a M.Sc. degree from the University of California Berkeley. Her research activities focus primarily on optimization and machine learning techniques for large complex networks.

Mor Kaspi is a Postdoctoral researcher in the Urban Transport Systems Laboratory (LUTS) at EPFL. He holds a B.Sc., a M.Sc. and a Ph.D. in Industrial Engineering from Tel‐Aviv University. His research interests are in operations management of transportation and logistics systems, focusing on modeling and optimization techniques.

Nikolas Geroliminis is an Associate Professor at EPFL and the head of the Urban Transport Systems Laboratory (LUTS). Before joining EPFL he was an Assistant Professor at the Department of Civil Engineering at the University of Minnesota. He has a diploma in Civil Engineering from the National Technical University of Athens (NTUA) and a M.Sc. and Ph.D. from University of California Berkeley.

2. An algorithm for general infinite horizon lot sizing with deterministic demand, Milan Horniaček
Are you interested in permanent optimization and adaptation of inventory levels in your firm? Do you always want to know what are optimal purchases of inputs for several calendar periods ahead? Do you want to know which recommendations will change when future conditions (prices of inputs, cost of holding inventories, ordering cost) change? The algorithm described in my paper will give you answers to all these questions. It gradually computes optimal purchases of input for each period, indicating which result is the final one (that will not change as a consequence of optimization of inventory levels for the following periods).

Milan Horniaček: Game theory, microeconomics, and operations research are areas of my teaching and research. I earned habilitation from Charles University in Prague. I have lectured at Charles University, Swedish School of Economics and Business Administration in Helsinki, Czech Technical University in Prague, Slovak University of Technology and Comenius University in Bratislava.

3. Decision support systems for district heating network development and energy production optimization, Matteo Pozzi
Optimization of energy production and distribution is an essential business issue for companies operating District Heating networks, given the significant investments necessary to develop the network and high operating costs to run the corresponding energy production plants. Technical, economic and regulatory constraints pose significant challenges that increase the value of adopting optimization tools to deal with this increasing complexity. The poster describes two innovative solutions that support leading multi‐utility companies in Italy in strategic, tactical and operational decision making processes that have been object of increasing attention by national and international players.

Matteo Pozzi holds M.Sci. in Physics with over 15 years of management consulting practice in Italy and abroad. Since 2009 is CEO of Optit, spin‐off company of the University of Bologna specialized in the development of DSSs based on state‐of‐the‐art Operations Research. Selected projects have been finalist at EURO Award and Wagner Prize.

4. Evaluation of MCDA methods, Chergui Zhor; Moncef Abbas
The introduction of reference points concept in the procedures of Multi‐criteria Methods dates back to the early seventies. This new idea was quickly implemented by some researchers of this time. In the same way, Paul Yoon and Ching Lai Hwang gave rise to the TOPSIS method in the early eighties. It is essentially based on some principles of the Multiobjective Programming. Some years later an improved version of this method, called the revised TOPSIS, was proposed by Deng and al. In this paper, we present a personal contribution to the previous established research works; it consists firstly in showing some weaknesses in the performance of the revised TOPSIS. On this basis, an improvement of this method (new proposition) will be introduced. Otherwise, through a computer program developed on Matlab, a comparative study between the revised TOPSIS and the improved one is presented (a statistical study based mainly on randomly generated instances), its purpose is to show the effectiveness of the new method.

Chergui Zhor, is Assistant Lecturer of Operation Research, Mathematics, Algorithmics, Software Programming and Multicriteria Decision Aiding at the Faculty of Mechanics, Ecole Supérieure Nationale de Technologie (ENST), since March 2011. A member of MCDA & OR laboratory and PhD student at Operational Research department, Université de Science et de la Technologie (USTHB), Algeria. She is the author of 2 scientific publications and 6 conferences which mainly focus on multicriteria decision aiding, multi‐objective optimization.


5. a platform for organizing challenges for solving optimization problems Szymon Wasik, Maciej Antczak, Jan Badura, Artur Laskowski, Tomasz Sternal is a cloud platform to host programming challenges for solving optimization problems originating from industry or science, by inviting researchers from all over the world to participate in programming competitions. Each researcher submits the source code of the problem's solution, then the system compiles it, executes in a homogeneous run‐time environment and evaluates using the predefined set of test cases. The evaluation result is presented using the live ranking. The environment supports almost any programming language, as well as Linux binaries, CMake packages, and linear programming solvers. We have already verified by judging over 1000 solutions.

Szymon Wasik is an Assistant Professor of Computer Science at the Poznan University of Technology. My research interests include algorithms design, modelling, simulation, and crowdsourcing. I also gained a lot of experience of implementing on‐line judge systems by organising programming contests and, currently, I use it to develop platform.

6. Wind power forecasting into the Brazilian electrical system to operational dispatch, Soraida Aguilar, Reinaldo Castro Souza
Different methodologies have been used to forecast wind energy for operational planning purposes. This work presents a review of the different approaches employed to forecast wind power generation in the Brazilian electrical system context. Some of these techniques were applied to forecast wind generation 24‐hour‐ahead for an hourly dataset from a Brazilian wind farm.

Soraida Aguilar received the Ph.D. degree in electrical engineer from Pontifical Catholic University of Rio de Janeiro (PUC‐Rio), Brazil, in 2015. She is currently a post‐doctoral research at PUC‐Rio.

Reinaldo Castro received the Ph.D. degree in statistics form Warwick University, Coventry, England, in1979. He is currently a professor at PUC‐Rio.

7. Product launch modelling and optimisation on the cardiology product market in a CEE country, Rossen Kazakov, Penka Petrova
A product market computer model was developed for a regional player on the CEE medicines market using resource dynamics modelling methodology, and then transformed to a business simulator which enabled the forecasting of alternative product growth trajectories related to different marketing resource configurations, considering the market as an interconnected non‐linear dynamic system including patient flows, doctors’ behavior, competitor reactions, ethical marketing and constantly changing P&R regulation. Using virtual performance experimentation by manipulating key factors and variables, what‐if product launch potential scenarios were explored. As a result, different actions were compared and optimal performance paths were mapped for selecting the most robust product launch design.

Penka Petrova, Manager at Strateggo Ltd., Chair of the World Business Web Association, has obtained LLM and Master in Finance degrees and has specialized in system dynamics modelling and simulation at UPC, Spain. Mrs Petrova has experience in the financial and banking field and participates in conferences of the System Dynamics Society, Strategic Management Society and the
OR Society. Member of the Swiss Operations Research Society.

Rossen Kazakov, Manager at Strateggo Ltd., has obtained MBA and Master in Economics and has experience in the pharmaceutical industry. Mr Kazakov is a part‐time PhD researcher at the Strathclyde University, and participates in conferences of the System Dynamics Society, Strategic Management Society and the OR Society. Member of the OR Society.

8. GIS application in design of mobile post office service routes – Hong Kong case study, Karolina Glowacka
This poster considers a mobile post office service in Hong Kong, predominantly serving elderly individuals, who cannot easily access a regular post office location. Geographic Information System is used to redesign mobile post office routes in order to increase the service’s reach while minimizing total travel distance. Solutions are presented for various values of model parameters, including relative importance of population reach versus distance to the nearest post office, number of mobile post office sites, and number of depots. The methodology described herein can be also applied to other services, such as mobile healthcare clinics and mobile libraries.

Karolina Glowacka is an Assistant Professor in Supply Chain and Information Management at Hang Seng Management College in Hong Kong. She received her PhD in Operations Research from University of Pittsburgh. She focuses on applied research in the areas of healthcare management and logistics.

9. Who do you think you are? Exploring the identity of OR professionals, Frances O’Brien
Who do you think you are? When you talk to other people about what you do, how do you describe your role? Do you see yourself as an OR professional with a strong sense of professional identity? Is your sense of identity is linked to the type of work you do and the experiences you've had throughout your life? If you haven't thought about it much, this poster will explore factors that influence the identity of OR professionals.

Frances O’Brien is an Associate Professor in OR/MS at Warwick Business School. She teaches business modelling and strategic development courses to both undergraduate and MBA students. Her research interests include the development and use of tools to support strategy development, in particular scenario planning and visioning methodologies.

10. Blending systems thinking approaches for organisational analysis: reviewing child protection in England, David C Lane
This work concerns the innovative use of a blend of systems thinking ideas in the 'Munro Review of Child Protection', a high-profile examination of child protection activities in England, conducted for the Department for Education. Three specific contributions that systems thinking made to the Review are described. First, the systems-based analysis and visualisation of how a 'compliance culture' had grown up. Second the creation of a large, complex systems map of current operations and the effects of past policies on them. Third, how the map gave shape to the range of issues the Review addressed and acted as an organising framework for the systemically coherent set of recommendations made. The poster ends with the main implementation steps taken so far to create a child protection system with the critically reflective properties of a learning organisation, and methodological reflections on the benefits of systems thinking to support organisational analysis.

David C Lane is Professor of Business Informatics at Henley Business School. He specialises in strategic analysis using system dynamics and systems thinking. He works interactively with groups of senior managers to help them express their ideas in the form of a system dynamics‐based simulation model. The model and the facilitated process help managers to think through the long‐term policies of their organisation.

11. Measures of personality and financial knowledge in consumer credit risk applications, Galina Andreeva
It is reasonable to suggest that borrowers’ personality is at least partially responsible for repayment performance, yet research on this topic is limited. This poster explores a potential value of personality measures in risk assessment and for Loss Given Default modelling. It will describe the preliminary results from the project based on a unique Brazilian dataset that combines traditional loan characteristics with survey data that includes measures of borrowers’ attitudes towards risk‐taking and their financial knowledge. The poster is targeted at a potential industry collaborator and data provider.

Galina Andreeva is a Senior Lecturer in Management Science at The University of Edinburgh Business School and a member of Credit Research Centre. Her research interests involve risk assessment in consumer credit and SME lending, discrimination in consumer credit markets, financial exclusion, financial integration in Europe, construction of cross‐country scoring models, profitability modelling.