Research

Due to increasing population in metropolitan areas, intelligent planning and control of urban mobility and freight transportation is becoming more important. Traditionally, research in the area of city logistics has investigated the trade-off between efficiency and reliability of urban deliveries. Extending well-known city logistics approaches, the Research Group Business Analytics focuses on the preparation and aggregation of large databases for dynamic and stochastic methods of transportation planning and control, e.g., optimal path-finding and vehicle routing. We follow a holistic approach stretching from data collection via data aggregation with data mining to extended approaches of applied operations research. We are also working on better planning and control of mobility and transportation services such as car and bike sharing services.

To check out our research in Green Logistics, watch the Flash Talk. 

 

 

Research Focus Area Simulation-Based Analytics (SimBA) – Smart Decision-Making for Challenging Applications

We live in a world full of dynamic and heterogeneous socio-economical and socio-ecological challenges. We want to address these challenges with simulation-based analytics (SimBA). SimBA is about data-driven, smart decision making including causal and expert information. The starting point is data, which can be derived from large operational databases. SimBA’s approach is that simulation of socio-economical and socio-ecological relationships and systems provide synthetic data. Simulation focuses on microscopic, agent-based models, but will include also other modeling approaches. The created data and information needs to be transformed into information models, which describe the complex relationships compactly. This information is considered in anticipatory decision models, which allow for making smart decisions. The impact of these decisions can be evaluated by simulation models again, and the whole process can be refined until the trade-off between economical and ecological goals has been explored sufficiently.

The following projects have been funded in this area:

For more details, please click on the particular project.

ExplAIn-TrAIn-Plan

© ÖBB / Harald Eisenberger

This project builds upon the outcomes of two previously completed projects: GreenTrAIn, which focused on green railway planning, and VIPES, which addressed the reliability of railway operations. Leveraging the expertise and methodologies developed in these projects, the current follow-up project aims to advance multi-objective railway resource planning for heterogeneous locomotive fleets, balancing energy efficiency, cost, and operational reliability.

To address the inherent complexity and scale of the problem, the project will develop advanced meta-heuristic optimization algorithms. A key innovation is the implementation of a simheuristic framework—an AI-driven integration of meta-heuristic optimization with agent-based simulation—to incorporate real-world uncertainties into the planning process.

The research will be grounded in real-world data, including infrastructure data, historical timetables, energy consumption, and delay data. This data will will be used for energy consumption and reliability predictions in optimization and simulation.

The consortium is led by dwh GmbH and consists of ÖBB Produktion GmbH, the University of Vienna (UW), the Technical University of Vienna (TUW) and the Technical University of Dresden (TUD). ÖBB will provide the required data, UW will focus on energy consumption modeling for heterogeneous locomotive fleet and multi-objective circulation planning, UW and TUW will collaborate on AI-based integration between optimization and simulation, and TUD will work on the integrated optimization and simulation of train drivers' schedules.

Cooperation partners

The Austrian Research Promotion Agency (FFG) is the central national funding organization and strengthens Austria's innovative capacity. ExplAIn-TrAIn-Plan (FFG project number 923467) is funded by the Federal Ministry for Climate Protection, Environment, Energy, Mobility, Innovation and Technology (BMK) as part of the “AI Ecosystems 2024: AI for Tech & AI for Green" call for proposals. www.ffg.at

Dynamic Tester Routing

Picture: www.meinbezirk.at

• Investigated by: Jan Fabian Ehmke, Marlin Ulmer, Niki Popper, Peiman Ghasemi

In the COVID-19 pandemic, we saw that rapid and efficient testing can effectively slow down the spread of a virus. Based on the idea of a mobile fleet of testers, we want to investigate in this research project how scarce resources can be used sensibly in a dynamic environment by planning ahead. For this purpose, modern methods and tools from agent-based simulation, data analysis and dynamic vehicle routing are used to examine the use of mobile tester fleets for COVID-19 cases. Abstracting from the COVID-19 example, we also want to analyze similar problems. When is the interaction of predictive planning in combination with machine learning helpful? How can detailed simulations improve forecasts? And how does one computationally feasibly integrate information from highly complex simulations into predictive planning? Examples here extend far beyond COVID-19, e.g., to demand forecasting for delivery services.

Participants in the project are Univ.-Prof. Jan Fabian Ehmke from the University of Vienna, who will focus on demand modeling. In addition, there is the expertise of Univ.-Prof. Dr. Marlin Ulmer, University of Magdeburg, who will be responsible for learning procedures related to predictive vehicle routing. Evaluation and demand generation will be done by Dr. Niki Popper (TU Wien), an expert in agent-based simulation.

Recent publications

Cooperation partners

This project is funded by The Austrian Science Fund (FWF): project number I 5908-G.

Modeling Complex Driver and Customer Dissatisfaction in Routing and Scheduling Problems

Picture from Oliver Hale auf Unsplash

• Investigated by: Tom Bormann, Jan Fabian Ehmke

In many routing models, customer dissatisfaction is simplified as a deviation from hard or soft time windows, overlooking the complexity of real-world preferences regarding when a service should start. However, in practice, customers often struggle to express their preferences precisely. This research addresses these limitations by modeling customer dissatisfaction as a linear piecewise function and introducing a natural language-based approach to capture vague or differentiated preferences. In doing so, it aims to better reflect individual expectations while also integrating driver dissatisfaction into the routing decisions. The project contributes to the intersection of survey methodology and routing and scheduling, with the goal of gaining insights into how realistic preference modeling can improve service planning while balancing the often conflicting objectives of customer and driver satisfaction.

Sensor-Based Waste Management

Picture from Burkhard Kaufhold auf Unsplash

 

• Investigated by: Jan Fabian Ehmke, Remy Spliet, Sanne Woehlk

Waste management is a challenging and costly business. On the one hand, we want to avoid overfilling of waste bins to keep our cities clean. On the other hand, a high frequency of visiting waste bins is very costly. In this project, we investigate the impact of sensors that provide information in waste management optimization. Based on stylized and real-world instances, we try to investigate where and when sensors make the most sense.

Recommendation Systems for Complex Travel Bundles

• Investigated by: Omar Aitoulghazi, Jan Fabian Ehmke, Rudolf Vetschera

In an era where traveling individuals seek seamless and personalized journeys, bundling travel services has become a challenge for several domain experts, whose tasks evolve around analyzing customers’ preferences, identifying partnerships, and managing profit sharing. Due to the continuous increase of customers, travel services, and their providers, curating personalized bundles manually has become quite complex. Nowadays, many travelers find themselves in front of an overwhelming number of options to choose from and manage in different booking platforms. This research focuses on automatically generating travel bundles inside one integrated platform by exploiting advancements in machine learning, such as recommender systems approaches. We expand the utility of these techniques to automatically learn customers’ preferred traveling experience attributes and how they change over time.

Research Group Prescriptive Business Analytics -- FWF Project: Using Machine Learning in BPC for Vehicle Routing Problems

• Investigated by: Juan Pablo Cantillana,  Christian Tilk

The planning and execution of deliveries represents a significant share of the costs in the area of distribution logistics and there is a great interest in exploiting the optimization potential in this area as best as possible. Therefore, route planning problems are also one of the central issues of Operations Research (OR). The use of optimization methods for route planning promises great potential for cost savings through the higher quality of the planning results as well as through the automation and acceleration of the planning process. There are numerous variants of route planning problems based on the different real-world requirements that transport service providers face. In most route planning problems, the task is finding the cheapest set of routes for a given fleet of vehicles such that a given set of orders is fulfilled while taking various side-constraints into account.
The most powerful exact route planning algorithms are based on branch-cut-and-price (BCP), which in turn is based on column generation techniques. Here, a master problem is responsible for the best selection of the available routes, while one or more so-called pricing problems successively generate new routes. The planned project is intended to generate new insights into the use of machine learning (ML) in BCP to solve route planning problems. The key question is how ML can be used in OR algorithms, because optimization problems are very different from most problems successfully solved by ML.

Cooperation partners

This project is funded by The Austrian Science Fund (FWF): project number P 37074.

Customer Acceptance in Attended Home Deliveries

• Investigated by: Charlotte Köhler, Ann Melissa Campbell, Catherine Cleophas, Jan Fabian Ehmke

Home delivery services require the attendance of the customer during delivery. Hence, retailers and customers mutually agree on a delivery time window in the booking process. However, when a customer requests a time window, it is not clear how much accepting the ongoing request significantly reduces the availability of time windows for future customers. We explore using historical order data to manage scarce delivery capacities efficiently and flexibly. 

Recent publications

• Flexible time window management for attended home deliveries

Also see

• efulfillmentresearch.org

Collaborative Transportation for Attended Home Deliveries

• Investigated by: Steffen Elting, Jan Ehmke, Margaretha Gansterer

Increasing congestion and concerns about harmful emissions in urban areas may soon motivate politics to install a collaboration platform for freight carriers to reduce the amount of delivery vehicles within cities. Moreover, by exchanging time window-based customer requests, carriers can gain additional profit. Combining methods of Operations Research and Game Theory, this project will investigate the interaction of dynamic customer acceptance and horizontal, auction-based carrier collaboration. The research goals are (1) to quantify the effect of dynamism in an auction-based horizontal carrier collaboration, (2) to determine an optimal methodological combination of delivery time slot management on the one hand and a combinatorial auction mechanism on the other hand, and finally (3) to investigate potential for strategic carrier behavior in both of these approaches.

Dynamic Vehicle Routing with Stochastic Customer Requests

• Investigated by Ninja Söffker, Marlin Ulmer, Dirk Mattfeld 

Many service providers offer customers to serve their request on the same day. In the case of a pickup or service request, the new customer can be integrated in the currently conducted vehicle routes. The service provider therefore needs to decide both about the acceptance or rejection of the new request as well as about the modification of the current routes while aiming on maximizing the number of customers or profit generated over each day of service. Our aim is to learn how to make these decisions in anticipation of potential future customer requests. To this end, we investigate an approximate dynamic programming approach with an adaptive state space partitioning.

Recent publications

• Adaptive State Space Partitioning for Dynamic Decision Processes 

• Stochastic Dynamic Vehicle Routing in the Light of Prescriptive Analytics: A Review