Research

Research lines of Michael Römer: decisions under uncertainty, health care analytics, network models and decision diagrams, language models for optimization, environment and sustainability, logistics.

My research sits at the intersection of Operations Research and Machine Learning. I develop decision support methods for planning problems in which uncertainty matters, and apply them in health care, logistics and environmental decision-making. A common methodological thread is the use of structured models, such as networks and decision diagrams, that are expressive enough for real-world problems and at the same time amenable to learning. Many papers belong to more than one of the lines below; papers are listed newest first. For funded projects, see Projects.

Decisions under uncertainty

Many planning decisions have to be made before demand, arrivals or durations are known. I work on two-stage, robust and data-driven approaches that account for this uncertainty, on quantifying the value of information, and on decision-focused learning, where predictive models are trained with respect to the decisions they inform.

  • Winkelmann, Ulrich, Römer, Langrock & Jahnke (2026). Managing inventories for perishable e-groceries: The value of probabilistic information. PLOS One. doi
  • Witteborg, Wetzel & Römer (2026). An anticipatory multi-stage MIP approach for the Integrated Healthcare Timetabling Problem. Operations Research, Data Analytics and Logistics, in press. doi
  • Römer & Bodur (2026). Learning-Guided State Partitioning for Decision Diagram Relaxations in Two-Stage Robust Optimization. Data Science meets Optimisation workshop @ IJCAI 2026.
  • Hagemann, Porrmann & Römer (2025). Multi-activity shift scheduling under uncertainty: The value of shift flexibility. European Journal of Operational Research. doi
  • Witteborg, Borgstedt, Rehberg, Wetzel & Römer (2025). Supporting ICU Admission Scheduling under Uncertainty. HICSS 2025 (Best Paper nomination). doi
  • Luhede, Yaqine, Bahmanbijari, Römer & Upmann (2024). The Value of Information in Water Quality Monitoring and Management. Ecological Economics. doi
  • Römer, Hagemann & Porrmann (2023). Predict, Tune and Optimize for Data-Driven Shift Scheduling with Uncertain Demands. LION 17. doi

Network models and decision diagrams

State-expanded networks encode complex rules, for example for shifts and rosters, in the structure of a graph, so that the remaining optimization model stays simple. Decision diagrams provide relaxations and bounds for discrete optimization. My work develops such models and uses machine learning to make them smaller and their bounds stronger.

  • Rudich, Cappart, López-Ibáñez, Römer & Rousseau (2026). An Exact Framework for Solving the Space-Time Dependent TSP. INFORMS Journal on Computing. doi
  • Römer (2024). Block-Based State-Expanded Network Models for Multi-Activity Shift Scheduling. Journal of Scheduling. doi
  • Nafar & Römer (2024). Using Clustering to Strengthen Decision Diagram Bounds for Discrete Optimization. AAAI 2024. doi
  • Nafar & Römer (2024). Strengthening Relaxed Decision Diagrams for Maximum Independent Set Problem: Novel Variable Ordering and Merge Heuristics. CP 2024. doi
  • Porrmann & Römer (2021). Learning to Reduce State-Expanded Networks for Multi-activity Shift Scheduling. CPAIOR 2021. doi

Health care analytics

Hospitals have to plan beds, staff and treatments while patient arrivals and lengths of stay are uncertain. Together with hospital partners and colleagues from medicine, I develop simulation-, learning- and optimization-based decision support for such problems: forecasting and managing intensive care occupancy, scheduling ICU admissions, scheduling clinical pathways, and rostering staff. I have also worked with clinicians and computer scientists on algorithm-based rehabilitation after cochlear implantation.

  • Witteborg, Wetzel & Römer (2026). An anticipatory multi-stage MIP approach for the Integrated Healthcare Timetabling Problem. Operations Research, Data Analytics and Logistics, in press. doi
  • Hagemann, Porrmann & Römer (2025). Multi-activity shift scheduling under uncertainty: The value of shift flexibility. European Journal of Operational Research. doi
  • Witteborg, Borgstedt, Rehberg, Wetzel & Römer (2025). Supporting ICU Admission Scheduling under Uncertainty. HICSS 2025 (Best Paper nomination). doi
  • Eichler, …, Römer, …, Todt (2022). Algorithm-Based Hearing and Speech Therapy Rehabilitation after Cochlear Implantation. Brain Sciences. doi
  • Witteborg, Borgstedt, Günther, Jansen & Römer (2022). Monte-Carlo Simulation Based on Patient-Individual Distributions for Supporting Intensive Care Occupancy Management. HICSS 2022. doi
  • Schwarz, Römer & Mellouli (2019). A data-driven hierarchical MILP approach for scheduling clinical pathways: A real-world case study from a German university hospital. Business Research. doi

I organized the 2024 annual meeting of the GOR working group “Health Care Management” in Bielefeld and won first place in the Second International Nurse Rostering Competition (INRC-II) in 2015.

Language models for optimization

Formulating an optimization model is often the bottleneck for using optimization in practice. In the project OptimizationChat we investigate how large language models can help to formalize and solve industrial planning problems through natural-language interaction. Related work uses language models to discover heuristics for combinatorial optimization.

  • Hottung, Berto, Hua, …, Römer, …, Tierney (2025). VRPAgent: LLM-Driven Discovery of Heuristic Operators for Vehicle Routing Problems. Preprint. arXiv

Environment and sustainability

Environmental decisions, such as where to monitor, what to protect and how to weigh ecological against economic goals, are made under substantial uncertainty and with limited budgets. Together with environmental economists and marine biodiversity researchers, I study the value of information in environmental monitoring, trade-off analysis for environmental decisions, and, in a DFG project, decision-focused learning for planning marine protected areas.

  • Post, Römer & Khanjani-Shiraz (2025). Strengthening Environmental Trade-off Analysis using Socio-economic Weight Restrictions. SSRN working paper. doi
  • Luhede, Yaqine, Bahmanbijari, Römer & Upmann (2024). The Value of Information in Water Quality Monitoring and Management. Ecological Economics. doi
  • Valipour, Babapour-Azar, Nourollahi, Khanjani-Shiraz & Römer (2024). Risk-driven optimal scheduling of renewable-oriented energy hub under demand response program and energy storages. Sustainable Cities and Society. doi

Logistics and supply chains

Many of my projects start from operational problems in logistics and retail: service network design, inventory routing, storage assignment and inventory management in e-grocery fulfilment, truck loading, and most recently large-scale supply network optimization in the automotive industry. Logistics also extends into space: a time-space network approach for delivery scheduling won first place in the “Delivery Scheduling” category of the ESA/GECCO Space Optimisation Competition 2022, and our exact framework for the space-time dependent TSP (INFORMS Journal on Computing) solves asteroid routing problems, in which a spacecraft visits a sequence of asteroids (Space.com).

  • Rudich, Cappart, López-Ibáñez, Römer & Rousseau (2026). An Exact Framework for Solving the Space-Time Dependent TSP. INFORMS Journal on Computing. doi
  • Winkelmann, Ulrich, Römer, Langrock & Jahnke (2026). Managing inventories for perishable e-groceries: The value of probabilistic information. PLOS One. doi
  • Winkelmann, Tolkmitt, Ulrich & Römer (2024). Integrated storage assignment for an e-grocery fulfilment centre: Accounting for day-of-week demand patterns. Flexible Services and Manufacturing Journal. doi
  • Neumann-Saavedra, Crainic, Gendron, Mattfeld & Römer (2020). Integrating Resource Management in Service Network Design for Bike-Sharing Systems. Transportation Science. doi
  • Schulte, Römer & Tierney (2020). Cyclical Inventory Routing with Unsplittable Pick-Up and Deliveries. ICCL 2020. doi

Team

Doctoral researchers

  • Since 2026 Hermann Gerzen: Large language models for optimization modeling
  • Since 2026 Johannes Hejny: Explainable optimization
  • Since 2026 Paul Eckert: Large-scale supply network optimization (external, in cooperation with industry)
  • Since 2024 Nele Bertling: Decision-focused learning for planning marine protected areas
  • Since 2021 Henning Witteborg: Data science and decision analytics for health care processes

Alumni

  • 2025 Dr. Jakob Schulte: Decision analytics for planning and scheduling in logistics
  • 2025 Dr. Mohsen Nafar: Bounds for decision-diagram-based optimization
  • 2024 Till Porrmann: Learning-enhanced network models for shift scheduling under uncertainty

Competitions

  • 2022 1st place, ESA/GECCO Space Optimisation Competition, “Delivery Scheduling”
  • 2016 1st place, ACP Torpedo Scheduling Challenge
  • 2015 1st place, Second International Nurse Rostering Competition (INRC-II)