When: Monday, 24 April 2023, 09:00-17:00.
Where: Industriens Hus, H. C. Andersens Blvd. 18, 1553 København, Denmark.
Applications of Optimization 2023 (AOO 2023), our annual conference on applied mathematical optimization takes place on Monday, 24 April 2023 at Industriens Hus in Copenhagen (09:00-17:00).
Below you find the programme, followed by the abstracts of the conference talks.
The workshop program includes five talks from OR experts and practitioners, and there will be ample time for networking:
09:00 – 10:00 Registration and refreshments
10:00 – 10:10 Welcome and announcements
10:10 – 10:55 Elaine Chew, Professor of Engineering at King's College London
Music, Mathematics, Digital Therapeutics
11:00 – 11:45 Sander van Aken, Senior Operations Research Engineer at Flix SE
Developing an OR-based software product to support network planners at Flixbus
11:45 – 12:00 Sponsor pitches
12:00 – 13:00 Lunch
13:00 – 13:45 Kristian Edlund, Director of Data Science Incubation at LEGO
Optimizing Quality Assurance sample scheduling
13:50 – 14:35 Trine Krogh Boomsma, Associate Professor at University of Copenhagen
MILP sensitivity analysis for the objective function coefficients
14:35 – 14:45 Sponsor pitches
14:45 – 15:15 Coffee break
15:15 – 16:00 Tommy Clausen, Software Engineer and Operations Research Expert at cVation
Optimization in Airport Ground Handling
16:00 – 16:05 Closing
Speakers and Abstracts
Elaine Chew, Professor of Engineering at King's College London
Title: Music, Mathematics, Digital Therapeutics
Abstract: The formal, mathematical representation of musical structures underlies many industrial applications ranging from music information retrieval to music generation. Musical structures can refer to how notes are organised into compositions or how they are transformed into sound in performance. In this talk I will show how heartbeats can be used to constrain composed structures to create music reflecting different cardiac conditions, and how performed music structures can be linked to cardiac response for digital therapeutics. The presentation will be accompanied by demonstration at the piano.
Sander van Aken, Senior Operations Research Engineer at Flix SE
Title: Developing an OR-based software product to support network planners at Flixbus
Abstract: Flix is Europe's largest player in the long-distance bus transportation market, transporting tens of millions passengers in 40 different countries. Thorough network planning enables unlocking Flix's vision of providing smart, sustainable, and affordable travel for everyone. Our network planners create timetable, bus and driver schedules for lines in the European network running between 1 and 51h. Although they can rely on a wide range of high-quality data sources and supporting tools, it remains a very manual and experience-based process.
Our team's mission is to provide data-driven solutions based on operations research algorithms enabling network planners to build customer-oriented and profitable networks. We recently successfully released our first product focusing on bus and driver scheduling, using a hybrid algorithm capable of generating high-quality solutions. As such, network planners can unlock cost savings, and spend more of their time focusing on the timetable design. We will talk about the road towards this first release, the challenges we faced in developing the product and algorithm, and how we move ahead.
Kristian Edlund, Director of Data Science Incubation at LEGO
Title: Optimizing Quality Assurance sample scheduling
Abstract: The LEGO Group has strict tolerances and standards for the bricks we produce. This results in a lot of samples being selected for evaluation to ensure that we always stay within tolerances. While the samples are awaiting analysis, we continue the brick production for up to 24 hours before a result is ready. In case of the samples being rejected this leads to lost production and scrapping of materials. In this talk, I will dive into getting an impact from improving the sample evaluation schedule and all the pitfalls of going from a desktop analysis to getting it implemented in an operational setting.
Trine Krogh Boomsma, Associate Professor at University of Copenhagen
Title: MILP sensitivity analysis for the objective function coefficients
Abstract: The talk introduces a new approach to sensitivity analysis of the objective function coefficients in mixed-integer linear programming (MILP). We determine the maximal region of the coefficients for which the current solution remains optimal. The region is maximal in the sense that for variations beyond this region, the optimal solution changes. For variations in a single objective function coefficients, we show how to obtain the region by bi-objective mixed-integer linear programming. In particular, we prove that it suffices to determine the two extreme non-dominated points adjacent to the optimal solution of the MILP problem. The methodology extends to simultaneous changes to two or more coefficients by use of multi-objective analysis. It can likewise be extended to sensitivity analysis in multi-objective MILP. Computational results illustrate the applicability of the approach.
Tommy Clausen, Software Engineer and Operations Research Expert at cVation
Title: Optimization in Airport Ground Handling
Abstract: Airport ground handling is the process of providing services to aircraft when on the ground. Airports are highly dynamic venues where unscheduled changes frequently occur, which make ground handling operations challenging to plan.
In this talk, we discuss how OR and software development are combined to create software systems for operational planning of ground handling. We review some lessons learned and how software engineering practices may improve OR systems in general.
AOO 2023 is sponsored by Mosek, Ørsted, Sealytix, and Shoreline!