Making An Impact 
Practitioners’ stream

Making an Impact: the practitioners’ stream, Room Q106

 

Are you interested in applying OR in industry, business or public sector organisations? Then ‘Making an Impact’ at  EURO2019 is for you. During the Making An Impact Stream at EURO2019 you can:

  • Explore the challenges every practitioner faces, and share solutions 

  • Try out new techniques – are they any good for you?

  • See case studies showcasing important applications

  • Exchange ideas and expertise with people in similar fields

  • Meet leading academics and discover what they can do for you – and what you can do for them

  • Build your network amongst likeminded professionals

 

“I really enjoyed the Making an Impact stream”  “…a great initiative..”  “..helpful and inspiring..” “..worthwhile..”   (2018 participants)

OVERVIEW
Monday 24th June
  • Peer networking session – meet like-minded people and discuss shared challenges, in a format designed to be comfortable to extroverts and introverts alike

  • Lightning talks – a high-intensity session packed with 5-minute talks from a wide variety of professionals including  John Poppelaars (author of ‘OR at work’ blog), Elise de Rosario, Prof Michael O’Neill

  • Practitioner’s presentations on Machine Learning and Operations Research, covering challenges and opportunities

Tuesday 25th June
  • OR practitioners’ talks with an emphasis on practical details, achievements and impact

  • Roundtable panels with industry, business and leading academic experts, exploring the practical challenges of applied OR

Wednesday 26th June
  • Roundtable Panel: Can OR save the world? (a joint session with the Governance Analytics stream) – join leading experts in the application of OR to sustainable development and social improvement, to discuss how OR can help respond to the enormous challenges now facing mankind

  • Tutorials:

    • Optimizing using Python Learning from data viavisualization and storytelling

    • Workshop: Using scientific benchmarks from the real world and OR competitions to help develop OR skills

Monday 24th June:
MA: Speed networking and EURO working group on practice of OR (Chair:Ruth Kaufman)
One of the most important reasons for coming to a conference is to build your circle of professional acquaintance - with like-minded people who can encourage and inspire you, and with not-quite-so-likeminded people who can challenge and extend you, and with the many people who you can encourage, inspire, and challenge. However, it is not always easy to meet people when most of our time is spent sitting listening to presentations, especially for the shy people amongst us. The speed networking session is a supportive, managed way of meeting people you don’t know: a series of short, quick-fire sharing of professional information with other participants, together with an exchange of contact details if you meet somebody you want to keep in touch with. Latecomers will be admitted but may have to wait for a few minutes before they can join in. Bring business cards, or be ready to scan QR codes with your linkedin app, to keep in touch.

The meeting will also include a short presentation about the EURO Working Group on Practice of OR, aimed at supporting a pan-European network of practitioners.

“I really liked the setup”… “Great start to the conference” (comments from 2018 participants)

MB: Fast and Furious: lightning talks  (Chair: Ruth Kaufman) At a EURO conference there is an enormous amount of interesting material that any one person can only see a small fraction of. This session will help boost that fraction, with a wide variety of fast-moving and stimulating talks. Into this 90-minute session, we will squeeze no fewer than 12 – maybe more! – ‘lightning talks’ from speakers at every level of eminence and experience. Some will be giving a compressed version of their talk later in the conference, whilst others will be talking on something completely different.

For full list with abstracts and author details, go to [bottom of page].

“A really good format … especially liked the 5 minute talks because they focused on the key messages instead of going into too much” (2018 participant)

 

MC:   Operations Research Meets Machine Learning: highlights from the 2nd conference of the EWG-POR (Matteo Pozzi)

More than 100 OR practitioners and industrial representatives met in Bologna on 11-12th March 2019 for the 2nd conference of the EURO Working Group on Practice of Operations Research to exchange experiences and best practices, with a particular focus on the relationship between Operations Research and Machine Learning (and Data Science at large). What came out was the picture of a rapidly evolving business community, rooted in its core skills, but also open to new challenges with a focus to developing increasingly effective Decision Support Systems. This session will report on the event and its key findings, representing an additional chance to foster debate and confrontation about this very actual issue.

MD: From Data to Decisions. The challenges of the digitalization era (Raffaele Maccioni)
It is not always so easy to take the best decisions, both strategic, like to redesign the supply chain, or operative like to deliver goods at min costs, or to define prices and promotions. Even more, sometimes it is not so clear what "best" really means. This is also true for the solution that has been awarded by Informs as a finalist of the Franz Edelman Award. ACT Operations Research provides an inspiring view on what it takes to develop and implement “deep-analytical” solutions.

Tuesday 25th June: focus on practice and industry

TA: OR impact across industries (Jean André)

a.       Airport Ground Staff Scheduling: A Real-World Business Application
Staff scheduling and rostering involves a number of hierarchical subproblems including demand modeling, task generation, shift design, days-off scheduling, shift assignment and real-time dispatching. When solving highly constrained large-scale workforce planning problems it is usually not computationally practical to deal simultaneously with all these tasks. Real-world software solutions typically decompose the overall planning task into heuristically designed subproblems which then are tackled by a variety of suitable exact and heuristic methods. We present results from a major research and business project with Swissport International, the largest ground handling company worldwide, which provides services for 850 customers and 265 million passengers a year, with a workforce of 68 000 personnel at 315 airports. During a long-term strategic cooperation, a high-performance software for automated staff scheduling has been developed, which is able to solve the complex large-scale rostering problems in Swissport's airport operations. The methodology comprises a broad range of optimization techniques including preprocessing, decomposition and relaxation approaches, mixed-integer programming models, and various heuristic procedures. Operational deployment started at Zurich Airport and is continually being extended to other airports and customers. Bottom line benefits include faster and more robust planning processes, improved roster quality, and substantial financial savings.

b.       Examples of mathematical decision support models for operations management and planning
We present examples on how development of mathematical decision support models and their implementations have improved operations management and planning in the Finnish state-owned railway operator VR. Our examples’ application domains include personnel planning, locomotive allocation, and maintenance scheduling. The developed solutions utilize, for example, mixed integer linear optimization, constraint programming, combinations of heuristic solution algorithms, and Monte Carlo simulation.

 

TB: OR impact on societal problems: (Chair:Michele Quattrone)

a.          Extending Capture-Recapture to Estimate Border Securitya.

To maintain secure borders it is vital to measure the effectiveness of border security measures. Capture-recapture techniques have been proposed, augmented by biometric identification to play the role of “tagging.” These models cannot deal with the confounding between deterrence and detection. We introduce a "two fluid model" which binarizes the effect of having been caught once. This permits estimation of the effectiveness of border security, by focusing on those identified persons who persist in border crossing, as probes of effectiveness. We explore the problems of (1) resolving the deterrence-detection confounding using non-linear optimization and (2) incorporating pseudo-Darwinian models, in which those persons who are repeatedly captured are modeled as belonging to the "low end" of a spectrum of "ability to avoid capture." We assess the theoretical analysis using simulation methods, and explore the path to application in realistic situations, where there are multiple paths by which to cross one or more borders. Research supported in part by the U.S. Department of Homeland Security Center of Excellence for Borders, Trade and Immigration.

b.            Justice in Time: Applying Operations Theory to Court Systems – The Case of the Jerusalem District Court


If there is some truth to the old principle that justice must not only be done, but must be seen to be done, then surely it must be seen to be done in a timely manner. Nevertheless, court congestion and delays—which threaten to undermine the justice system—became a global phenomenon with significant adverse implications on social welfare, economic development and civil rights. In this research, we report on the application of several operations management tools to tackle this problem and help reduce congestion in courts. In cooperation with the Jerusalem District Court in Israel, we have implemented the following changes in the judicial process: (1) Introduction of a new planning and scheduling phase; (2) Modification of the hearing scheduling policy from hearing-based FIFO policy to case-based FIFO policy; (3) Shortening the summations stage by encouraging oral summations. We evaluate the effect of these operational changes by performing a before-after econometric analysis. Our results demonstrate a 40% reduction in trial time (the treated part of the process), which translates into 18.87% reduction in the total life cycle of a case.

c.           Letting a hundred data scientists bloom: harnessing the power of cutting-edge data science to improve the outcomes of critically ill children
BACKGROUND: Advanced data science has been hailed as a game changer in delivering better health care. The UK NHS has many large datasets but these are messy, hard to interpret and the use of advanced data science techniques is still in its infancy. Operational Researchers could be key in meeting the challenge of how to mesh the highly technical world of data science with the realities of front-line clinical decision making. The UK Turing Institute for data science runs regular Data Study Groups (DSGs) where a hundred top data scientists from all fields gather for a week of intense work on real-world problems. Great Ormond Street Hospital (GOSH) has a unique high resolution vital sign dataset of 5500 children in intensive care and contributed a problem to the DSG for April 2019.
THE PROBLEM GIVEN TO THE DSG: When children are on life-support, their vital signs are monitored continuously. Clinical staff make treatment decisions informed by current snapshots of these signs. The challenge was to develop predictive algorithms that out-perform clinical decision-making for one high-risk decision faced for every patient: optimal timing for a trial of breathing without life-support.
THIS TALK: I am an Operational Researcher and will discuss how I worked closely with GOSH and others to conceive of, negotiate, plan, execute and interpret the hospital’s foray into new data science. I will discuss the skills needed, the challenges encountered and the preliminary results.

 

d.            Blockchain in food distribution: do you know where your food has been?
This paper aims to explain the implementation of blockchain technology in the production and supply chain delivery system for beef/cattle by a company based in the USA (Bytable Foods). One of the primary research questions is how blockchain can more accurately and transparently move goods through supply chains. This company, and paper, are at the leading edge of developing such systems in industry. Bytable Foods has various initiatives in development on the front line of blockchain. Bytable Foods is making food traceable and transparent through trustworthy data chains that track products from farm to fork using blockchain technology and internet of things. By creating traceable and transparent supply chains for food, consumers are given the information they need to make the right choice about the food they buy and the companies they support. For stakeholders in the food supply chain, having traceability and transparency builds better relationships with their customers, increases efficiency, and reduces the risk and cost of food recalls, fraud, and product loss. The technology and business are creating a case for fixing and transforming the world’s food system.
One case study of interest in the academic and industry community is related to the cattle industry. Bytable was founded in Iowa and is based in Colorado while IBM is working on a similar application called Beefchain.

TC: Round table 1:  Making an impact across industry (Chair:Jean André)
Around this roundtable, the panel will talk about the key performance indicators (economical, environmental...) industrial companies are trying to optimize jointly or separately (included or not in the models), the impact through examples on the operations to transform the way they work, the integration within their working environment and processes (user interactivity, HMI and others), the qualitative benefits such as team training, knowledge sharing... They will also cover the challenges of understanding the business needs and transform it iteratively into optimization models, the modeling effort to capture all the dynamic situations faced by operations, the difficulties of user acceptance vs the other tools they have used in the past (comparative effect) or home made tools or well marketed magic tools, how they are guaranteeing the actual use…

TD: Round table 2: Successful factors of industry/academics partnerships (Chair: Michele Quattrone)
Around this roundtable, the panel will discuss about the key ingredients that make an industry/academic partnership successful from an industrial and academic points of view. How do we initiate such partnerships? How do we track the evolution? Do we need to precisely define the roles in such partnerships ? What is expected from industrials by academics: data, real life topics, business expertises...? What is expected from academics by industrials: scientific excellence, modeling practice, algorithmic developments, hype technological check...? What are the final output of such collaboration expected: prototyped models, feasibility studies, Proof of concepts...? What are the issues that are regularly faced in such collaborations: misunderstandings, Intellectual Property,...?

 

 

 

Wednesday 26th June: panel,  tutorials and workshops

WA:  How can better governance save the world with the help of OR? a joint effort of the streams on Governance Analytics and Making an Impact, on the Practice of OR (Chair: Joaquim Gromicho)             

 

The United Nations just published an alarming report https://www.un.org/sustainabledevelopment/blog/2019/05/nature-decline-unprecedented-report/ showing that mankind not only leads to an accelerated rate of extinction of species but is orchestrating its own extinction, unless a dramatic change occurs. This calls for effective global governance measures. Intentions are however not enough: for instance agreeing on an amount of emission reduction is not the same as knowing how to reach such objective. Resources are also unequally distributed: we produce more than enough food to feed mankind, nevertheless more than 10% of the world is starving. Careful planning may lead to bringing food to those that need it at low cost and on time. Means of mass transportation have a lower environmental impact but are difficult to plan to the full extent of their potential. Huge opportunities for OR to develop the instruments that effective governance may use to improve the condition of our planet. The panel brings together experts in governance, sustainability, energy, societal impact, mass transportation, defense against the forces of nature and on how to effectively feed the world. The panel includes recipient of past Frans Edelman awards for impact on profit, people and planet. The panel discussion is started by Cathal MacSwiney Brugha.   The panel members are:

  • Cathal MacSwiney Brugha, Emeritus Professor of Decision Analytics, University College Dublin, President, Analytics Society of Ireland.

  • Anna Nagurney, tutorial speaker at the EURO, https://www.euro2019dublin.com/tutorial-speakers, focuses her research on network systems from transportation and logistics with emphasis on sustainability and quality issues with applications ranging from pharmaceutical and blood supply chains to perishable food products and humanitarian logistics.

 

WB: Tutorial: Optimizing with Python (Leader:Joaquim Gromicho)
Academics in OR learned a multitude of programming and modeling languages, all with their strong and weak points. Nowadays general purpose languages exist and python seems to dominate on all domains of analytics, but how good is it for optimization? That is exactly what this tutorial will teach you, taking into account mathematical modelling and optimization and also the development of heuristics for hard and challenging problems.

WC: Tutorial: let the data do the talking (Leader: Jasper Lodder)
As Tom Davenport (known for proclaiming Data Science to be the most sexy job on the planet in 2012) says in https://www2.deloitte.com/insights/us/en/topics/analytics/data-driven-storytelling.html “Most people can’t understand the details of analytics, but they do want evidence of analysis and data. Stories that incorporate data and analytics are more convincing than those based on anecdotes or personal experience. Perhaps the most compelling stories of all are those that combine data and analytics, and a point of view or example that involves real people and organizations.” Jasper Lodder will take us through the available technologies and methodologies for effective analytics storytelling, all the way from the first contact with messy data to taking decisions based on optimization models.

WD: Workshop: how hands on helps OR with benchmarks and challenges (Chair: Alex Fleischer)
We love a good puzzle, what about you? And are you in for a challenge? We believe that puzzles and challenges are great instruments to develop your OR skills. We will share our experiences and convince you that having fun makes you a better OR professional! Kaggle is famous among the data scientists and prediction analytics professionals and has no comparable optimization counterpart, but there are many options to challenge and improve your optimization skills. We report on great puzzles, benchmarks and challenges and on our experience on using challenges as an instrument to teach OR.

 

 

Fast and Furious lightning talks

List of speakers and titles (full details of each talk and speaker are below)

Elise de Rosario:  OR at your service

Michael O’Neill:  Business Analytics Capability, Organisational Value and Competitive Advantage

Irene Pluchinotta:  Being a decision analyst in collaborative decision-making processes: lessons learnt  

John Poppelaars:   Modelling at the speed of light?  

Paul Harper: OR Saves Lives!  

Mary Conlon:  From CT scanning to process modelling  - a radiographer's experience of modelling busyness   

Truls Flatberg: The value of simple models  

Ine Steenmans:  Behind-the-Scenes: current OR in the UK Government’s Industrial Strategy  

Georgio Sartor: The hotspot problem  

Tom Cnops:   Between optimization and the real world  

Marion Penn:  Using generic modelling to increase implementation: advantages and challenges  

Gavin Bell:  5 lessons in 5 Minutes for Success in Practice   

Jelke van Hoorn: Benchmarks from practice

Details

Ine Steenmans Behind-the-Scenes: current OR in the UK Government’s Industrial Strategy

People love the idea of systems thinking. OR analysts practise the principles of systems thinking. And governments are increasingly (re)appreciating the value of systems thinking. But when it comes to what it really looks like when a government uses OR tools for systems mapping on real policy problems, it is infrequent that we actually get access to the details ‘behind-the-scenes’. In this talk I’ll share a few views into ongoing work using systems mapping for the UK’s Industrial Strategy. Specifically, the UK has embraced the idea of ‘missions’ to inspire government and business to work across sectors and regions to achieve  bold and inspiring goals (inspired by Kennedy’s prompt to US industry and research to ‘get a man on the moon’!). OR tools are being used to make sense of what such a mission looks like in the world today.

Ine is a practice-based researcher in Futures and Analytics for policy at University College London (UCL). She looks at the practical ways OR tools and analytics are used to make sense of big, messy policy issues.

Tom Cnops   Between optimization and the real world

While modelling accuracy is always important, it is especially so when the model is not only used to describe but also to prescribe. One of the activities at Jeppesen is to help airlines create work schedules (rosters) for their crew. As such, the modelled constraints must exactly match what both crew and planners expect. The quality of the final product then not only depends on the quality of the optimization (“Solving the problem”) but also on the quality of the implementation (“Solving the right problem”).

 Although this is changing now, many airlines still use human intelligence in varying degrees to solve the rostering problem. As a result, rigorous definitions of the various constraints are usually absent and instead the process relies on the interpretation of the planners. This talk will discuss several examples where going from an obvious interpretation to a comprehensive rule is nontrivial.

Educated as a metallurgical engineer, Tom got into optimization while working on the supply chain of a steel mill. He now works on crew and operations management for Jeppesen.

Mary Conlon:  From CT scanning to process modelling  - a radiographer's experience of modelling busyness

Demand for diagnostic imaging services is increasing year on year, driven by Ireland’s growing, ailing and ageing population. Radiology workload, when defined in crude numbers of examinations completed, fails to capture staff workload and patient complexity.  As a clinical radiographer undertaking a PhD in healthcare modelling I have been in a perfect position to examine the use of operational research techniques, as communication and modelling tools. I have analysed radiology data pertaining to service demand and patient complexity factors (age, mobility and infectiousness) and incorporated it into discrete event simulation (DES) model of CT (Computed Tomography) services in a diagnostic imaging department. The validated simulation provides a decision support tool for managers and policy makers.  This presentation describes how simulation can facilitate data driven change of work practices and policies, allowing for better utilisation of limited resources; and shows some of the practical challenges encountered. Mary Conlon is a clinical radiographer (HSE) and 4th year PhD student (NUIG): Modelling the radiology service, creating a decision support tool and simulating alternative service delivery options.

 

Gavin Bell:  5 lessons in 5 Minutes for Success in Practice

OR in practice, I think, has one very simple goal - take a system or process, and make it work a little bit better. Over the past 20 years as a practitioner I have made many mistakes in trying to achieve this goal. Mostly these were not the result of classical modelling errors or a lack of model sophistication, but rather of something more mundane and simultaneously more difficult: misunderstanding the system and process, and what “making It work a little better” really means.  Fortunately - hopefully - I’ve learned something from these mistakes, and have boiled some of my experiences down into 5 lessons for a 5 minute talk: 1. Meet your customer’s capabilities; 2. Real world systems are compromises; 3. Always challenge wishful thinking; 4. Don’t let optimality trump usability; and 5. Academics are (mostly) your friends. Gavin is CEO of Optimeering, a developer of decision-support systems for the power industry. Gavin has a PhD from the University of Canterbury and over 20 years’ experience of OR in practice

Giorgio Sartor The Hotspot Problem:  peak and off-peak workload

In Air Traffic Management, a hotspot represents a region of the airspace in which too many airplanes are scheduled to fly through. The Hotspot Problem (HP) consists of determining new (possibly delayed) departure times for all flights to get a hotspot free schedule with minimum total delay.

The capacity of a region is related to the workload of air traffic controllers, and its duration: controllers can only sustain a heavy workload (i.e., several airplanes) for a short time. A light workload (i.e., only a few airplanes) can be sustained throughout the entire day.

Current methods measure workload over fixed windows of time, causing a phenomenon called "bunching": excess airplanes from a time-window are simply moved to the beginning of the next time-window.  This talk shows how a novel approach, sliding time-windows with variable width, both prevents bunching and differentiates between peak (small time-window) and off-peak (large time-window) restrictions.

Giorgio is a PostDoc Researcher at SINTEF, currently working on scheduling/dispatching/timetabling of trains and airplanes. He has a PhD in Applied Math from Singapore University of Technology and Design, and degrees in Automation Engineering from the University of Padua.

Irene Pluchinotta: Being a decision analyst in collaborative decision-making processes: lessons learnt

Collaborative decision-making is an intricate task for decision makers and decision analysts. Governments, public bodies and private-sector are beginning to involve stakeholders to a far greater extent than before. Actors with different values, points of view, rationales, and conflicting objectives are brought together in a collective process. Stakeholders involvement increases the need for effective decision-making processes and for structured stakeholder engagement activities.  How to handle a more inclusive decision-making process in multi-stakeholder settings? How to foster collaborative innovation in the case of decision makers operating in multi-organizational arrangements that tackle complex decision problems? The talk will share lessons learnt from series of facilitated workshops integrating OR methodologies and Design Theory. It will present an approach to formally support innovation in more collaborative, interactive and horizontal decision-making processes through the lens of real-life case studies.

 Irene is a researcher at King’s College London with a double PhD in environmental engineering and computer science. She uses OR methodologies to support collaborative decision-making processes in complex multi-stakeholder settings.

John Poppelaars   Modelling at the speed of light?

One of my visions as an OR practitioner is to start a meeting with a client, have some data and a vague description of a decision problem, and within hours, end the meeting with a fully-functioning prototype of a decision support system.  Impossible?  Perhaps. It is true that quite a bit of time is required to complete the path from data to a solution.  As a result, many decision makers shy away from OR as they need a solution fast.  The current reality is that we quickly get side-tracked by IT challenges such as coding, debugging, ETL nightmares and dealing with GUI development.  In this talk, I would like to show you that change is on its way!  Of course, not at the speed of light, but fast enough to make my vision become a new reality.

John Poppelaars is practice leader Advanced Analytics at BearingPoint and author of the OR at Work blog

Marion Penn: Using generic modelling to increase implementation: advantages and challenges

Generic modelling has the potential to significantly reduce the time and resources required to create models once the initial model has been developed.  This has the potential to increase the use and therefor also implementation of OR modelling in a variety of domains.  In healthcare a variety of organisations have similar structures, so there should be considerable scope for generic modelling to increase the impact of modelling.

This talk will explore the advantages of generic modelling with reference to an example from a healthcare context.  It will also consider the challenges of generic modelling and examples of steps that can be taken to address them.

Marion Penn works at the University of Southampton within the Wessex CALHRC (Centre for Applied Leadership in Health Research and Care), applying OR techniques to support research in healthcare contexts.

Paul Harper: OR Saves Lives!

Healthcare systems are stochastic in nature; that is they typically operate in an environment of uncertainty and variability, both at scale and within highly complex and connected networks.  Furthermore, many healthcare services are under significant pressure to deliver more with less.  OR methods can help healthcare providers move towards optimally configured services.  This is much safer than experimenting with changes to the system for real and seeing what happens.  Literally it can help save lives, for example in one major hospital our research completely redesigned the care for stroke patients, which resulted in a reduction in mortality rates by 60%.  In another hospital emergency department, our work helped to save the Health Board  £1.6m per year through optimised capacity planning.  In this talk I will provide an overview of some current and recent OR healthcare modelling projects at Cardiff, highlighting different OR methodologies and demonstrating impact.

 Paul Harper is Professor of OR in the School of Mathematics at Cardiff University, Deputy Head of School, and Director of the University’s Data Innovation Research Institute.  www.profpaulharper.com

Truls Flatberg: The value of simple models

In a highly complex and networked world, operations research models are also increasing in size and complexity. This trend makes models harder to understand for users, tending to decrease their confidence in the model and its results. In this talk I will focus on the value of simple models, illustrated by a real-world example from the aluminium industry. Even though being limited in scope, these simpler models present benefits in terms of securing user confidence and being computationally easy, allowing fast feedback in a distributed setting.

 Truls Flatberg is a researcher at SINTEF Industry working on applying operations research technology to create a better society.

Prof Michael O’Neill: Business Analytics Capability, Organisational Value and Competitive Advantage

Business Analytics makes the assumption that given a sufficient set of analytics capabilities exist within an organisation, the existence of these capabilities will result in the generation of organisational value and competitive advantage. Taken further, do enhanced capability levels lead to enhanced impact for organisations? Capability in this study is grounded in the four pillars of Governance, Culture, Technology and People from the Cosic, Shanks and Maynard capability framework. We set out to undertake the first empirical investigation to measure if there is a positive relationship between Business Analytics capability levels as defined by Cosic, Shanks and Maynard, and the generation of value and competitive advantage for organisations. Do enhanced capability levels lead to enhanced impact? A survey of 64 senior analytics professionals from 17 sectors showed a strong and statistically significant correlation between higher capability levels and the ability to generate enhanced organisational value and competitive advantage.

Michael O'Neill is Full Professor of Business Analytics in the UCD School of Business, and a founding Director of the UCD Natural Computing Research & Applications Group. http://people.ucd.ie/m.oneill/

Jelke van Hoorn: Benchmarks from practice
The benchmarks used to compare state-of-the-art algorithms are often random generated benchmarks. However, practical data does not only contain much more constraints and/or different objectives, also the structure of the data itself is often different. Wouldn’t be better to compare our algorithms on benchmark sets generated from examples from practice instead of the random generated instance we currently use? Jelke is an OR-engineer at ORTEC for over twelve years, currently working on 3D-binpacking algorithms. In the meantime he acquired a PhD at the VU Amsterdam.

Elise del Rosario: OR at your  Service

Are you providing OR services to an internal or an external client? As a manager or member of a  group of internal consultants, you deal  with such issues as establishing your credibility, sourcing and justifying your staff complement,  why the group has to be centralized or decentralized; prioritizing projects and whether or not to charge for services. As an external consultant however, your concern becomes more of defining and managing project expectations on scope and deliverables as you prepare your proposal.  In both cases however, it is important to have skills in communicating your findings and ensuring  that there is a project champion who sincerely wants an objective basis for decision-making, rather than one who just wants you to prove them right. This talk draws on many years of experience in both roles, to discuss these issues. Elise del Rosario was VP of OR in  San Miguel Corporation, She founded ORSP (Operations Research Society of the Philippines), headed the ORSP Committee on OR for Public Service (ORSP Corps) and was President of IFORS.

 

Are you a practitioner?

Become a member of the Work Group here.

Membership is Free