Announcement of new topics

Please note that topics are posted on a continuous basis.

Application for theses

Interested in one of our topics? Please use the application form!

Application Form

Current Theses (partially in German)

  • An AI for the game Quarto!

    2025/02/13

    Bachelor thesis

    Quarto! is a game that can be modelled as a graph and which has a compact action space while still being interesting and complex. The aim of the thesis is to develop an “AI” which is able to play Quarto!, as well as a GUI, helping to interact with the software tool. Based on the preferences of the student, the “AI” can be based on traditional algorithms (e.g., MiniMax) or machine learning (e.g., reinforcement learning).

    Interested in this topic? Please use the application form on our website!

  • Modelling approaches of the traveling salesperson problem

    2025/02/13

    Bachelor thesis

    The traveling salesperson problem (TSP) is a classical OR problem. A person, starting from a node, needs to visit a set of nodes before returning to the initial position. Hereby, the distance covered needs to be minimized. As the problem is known for years, many mixed integer linear programming formulations have been published over time. The purpose of the thesis is to conduct a systematic literature review, identify different modeling approaches, and explain the differences, pros, and cons of the single models.

    Interested in this topic? Please use the application form on our website!

  • Optimizing routes in e-commerce warehouses via reinforcement learning

    2025/02/13

    Master thesis

    Reducing unproductive walking times is an important planning problem in modern e-commerce picker-to-parts warehouses. In such facilities, the worker traverses the warehouse to collect all items of an order. Items of the same product are hereby stored at different positions within the warehouse, such that only a subset of storage positions must be visited. The so-called picker routing problem in scattered storage warehouses aims at a shortest tour through the warehouse, which allows the picker to collect a sufficient number of items per product before returning to a decentralized depot system.

    The aim of the thesis is to develop a reinforcement learning approach that optimizes the picker routing problem in mid-sized scattered storage warehouses. The famous taxi environment can be used as a starting point of the thesis.

    Interested in this topic? Please use the application form on our website!

  • Optimizing routes in warehouses via reinforcement learning

    2025/02/13

    Bachelor thesis

    Reducing unproductive walking times is an important planning problem in picker-to-parts warehouses. In such facilities, the worker traverses the warehouse to collect all items of an order. The so-called picker routing problem aims at a shortest tour through the warehouse, which allows the picker to visit all pick positions before returning to the central depot.

    The aim of the thesis is to develop a reinforcement learning approach that optimizes the picker routing problem in a small-sized warehouse. The procedure can, but does not have to, be adapted from the famous taxi environment.

    Interested in this topic? Please use the application form on our website!

  • Solving container relocation problems at intermodal container terminals

    2024/05/16

    Bachelor thesis

    Due to the growing global container trade, efficient handling of containers is essential. To optimize the performance of terminals, efficient container relocation (reshuffle) is crucial, especially in intermodal terminals where container relocation plays an important role. Therefore, this work aims to employ the beam search method (or branch-and-bound method) to achieve the minimum number of container relocations.

    Interested in this topic? Please use the application form on our website!

    Supervisor: Setareh Behzadi, M.Sc.

  • Container relocation problem: application of a reinforcement learning approach

    2024/05/16

    Bachelor thesis, Master thesis

    Efficient container relocation, or reshuffling, is important for terminal yard management, especially with the increasing global volume of containerized trade. To solve the container relocation problem in intermodal terminals, this work focuses on a simple yard structure and aims to investigate the application of reinforcement learning, especially the Q-learning method. The results, such as the relocation rate, will be assessed using a heuristic approach.

    * Previous knowledge of Python and reinforcement learning methods is mandatory.

    Interested in this topic? Please use the application form on our website!

    Supervisor: Setareh Behzadi, M.Sc.

  • Optimizing container operations at multimodal terminals: A literature review of machine/reinforcement learning methods

    2024/05/16

    Bachelor thesis

    The growing global container trade requires efficient handling and transportation of terminal containers to optimize the performance of inland container terminals and ports. As the demand for fast, efficient transshipment of terminal containers increases, innovative approaches are needed to improve measures such as task completion time, energy consumption, and overall operational efficiency. In multimodal terminals, the cranes generally serve the container ships, trucks, rail, and stacking areas. Unproductive movements of the cranes, for example in container relocation (reshuffling), should be minimized to improve the efficiency of the terminal. Intelligent methods such as machine/reinforcement learning can provide potential solutions. Therefore, this work aims to review the existing literature and develop innovative solutions for managing container relocation.

    Interested in this topic? Please use the application form on our website!

    Supervisor: Setareh Behzadi, M.Sc.

  • Benders Decomposition for robotized delivery on the last mile

    2023/05/24

    Master thesis

    The aim of the thesis is to understand the concept of Benders Decomposition and to apply it to the delivery of customer goods on the last mile, making use of the concept “vans and robots”. Hereby, the work of Alfandari et al. (2022) sould be understood and slightly extended.

    Literature: Alfandari, L.; Ljubić, I.; da Silva, M.D.M. (2022): A tailored Benders decomposition approach for last-mile delivery with autonomous robots. European Journal of Operational Research, 299(2), 510-525.

    Interested in this topic? Please use the application form on our website!

    Supervisor: Prof. Dr. Felix Weidinger

  • The Branch-and-Bound procedure for the classical Traveling Salesman Problem: Building a Demonstrator as an Open Educational Resource

    2023/05/24

    Bachelor thesis

    The Traveling Salesman Problem (TSP) as well as the Branch-and-Bound procedure (B&B) are both classics in Operations Research, tackled in many courses at different universities. The goal of this thesis is to develop a demonstrator which applies B&B to the TSP, using some simple bounds and branching schemes. The algorithm, hereby, needs to be visualized in a suitable manner, such that it can be used for educational purposes. The goal is to provide the demonstrator as an Open Educational Resource, ultimately, such that the outcome of the thesis can be used freely in any Operations Research course.

    Interested in this topic? Please use the application form on our website!

    Supervisor: Prof. Dr. Felix Weidinger

  • Development and benchmarking of a heuristic solution method for the picker routing problem in mixed shelves storage warehouses

    2023/05/10

    Bachelor thesis, Master thesis

    The steady growth of the e-commerce industry, especially fuled by the pandemic, puts increased pressure on various warehouse operations. One potential approach to increase order picking efficiency is to reduce picker walking distances. The unique feature in mixed shelves storage warehouses is that items to be picked can be located in multiple storage positions in the warehouse. This results in a multi-layered optimization problem: Suitable positions must be selected as well as the shortest route between them has to be found. The goal of this thesis is to develop a heuristic solution for the picker routing problem in mixed shelves storage warehouses, to implement it and to test it against existing methods.

    Interested in this topic? Please use the application form on our website!

    Supervisor: Constantin Wildt, M.Sc.

  • Optimization of randomized control rules for cleaning robots

    2022/03/29

    Bachelor thesis

    Decentralised acting cleaning robots usually operate on a static rule set. The aim of this work is to simulate different rule sets in different environments and to identify reasonable rules. Knowledge of implementation or the willingness to learn is a requirement for this.

    Interested in this topic? Please use the application form on our website!

    Supervisor: Prof. Dr. Felix Weidinger

Based on the topics that have already been worked on, you can gain an overview of relevant topic areas, to which you can refer to when developing your own topic proposals.

  • A literature review of machine/reinforcement learning methods for optimizing warehouse processes

    2023

    Bachelor thesis

    The purpose of this literature review is to provide an overview of machine/reinforcement learning methods that can be used to optimize warehouse processes. Various types of machine/reinforcement learning algorithms will be explored in this review, including those used for inventory management, item placement, and delivery. The challenges and limitations of using machine/reinforcement learning to optimize warehouses will also be discussed in order to suggest future research directions.

    Supervisor: Setareh Behzadi, M.Sc.

  • Discussion of several machine learning methods that can be used for predicting order picking times in warehouses

    2023

    Bachelor thesis

    This topic would involve discussion about several machine learning methods such as regression analysis, decision trees, neural networks and random forests methods. The focus would be on describing property of models and identifying which data would be used to train the models in warehouse.

    Supervisor: Setareh Behzadi, M.Sc.

  • Analysis of Regression Models for Optimizing Warehouse Picking Routing

    2023

    Bachelor thesis

    This study will examine a number of regression models for optimizing warehouse picking routing. Regression models including linear regression, multiple regression, and logistic regression will be compared and analyzed. This study will also identify factors affecting regression model accuracy in a warehouse, factors such as distance, time, and item location and so on. The results of this study will provide insights into the most effective regression models for optimizing warehouse picking routing.

    Supervisor: Setareh Behzadi, M.Sc.

  • Operations Research in the Aviation Industry: A Literature Review on Applications, Developments and Challenges

    2023

    Bachelor thesis

    The complex system of aviation involves numerous planning problems, which must be solved by different stakeholders. Examples include route, network, and crew planning by airlines or the allocation of logistics resources to aircraft and aircraft to gates by the airport. Challenges are, among others, the numerous constraints that must be considered in aviation as well as the coordination between the stakeholders. The aim of this paper is to identify and present (potential) areas of application of OR methods in the aviation industry through a systematic literature review.

    Supervisor: Constantin Wildt, M.Sc.

  • Analyse von Dokumentationen und Dokumentationszeiten

    2023

    Master thesis

    Supervisor: Prof. Dr. Felix Weidinger

  • Prescriptive Analytics (2017-2023)

    2023

    Bachelor thesis

    Supervisor: Prof. Dr. Felix Weidinger

  • Identifikation und Bewertung von Heuristiken für das Traveling Salesman Problem

    2023

    Bachelor thesis

    Das Traveling Salesman Problem (TSP) ist ein kombinatorisches Optimierungsproblem, welches in der Logistik unter anderem für die Planung von Lieferrouten eingesetzt wird. Aufgrund der Komplexität des Problems wird zur Lösung häufig auf Heuristiken zurückgegriffen. Ziel der Arbeit ist es, mittels einer Literaturrecherche vielversprechende TSP-Heuristiken zu identifizieren, diese anschließend zu implementieren und anhand von Rechenzeit und Lösungsqualität zu bewerten. Implementierungskenntnisse oder die Bereitschaft zur Einarbeitung sind hierbei eine Voraussetzung.

    Supervisors: Prof. Dr. Felix Weidinger, Constantin Wildt, M.Sc.

  • An exact approach for the picker routing problem in mixed shelves storage warehouses

    2022

    Master thesis

    Supervisor: Prof. Dr. Felix Weidinger

  • Entwicklung einer KI für das Spiel "Take it easy"

    2022

    Master thesis

    Supervisor: Prof. Dr. Felix Weidinger

  • Waveless Order Batching: Ein Deep Reinforcement Learning Ansatz

    2022

    Master thesis

    Supervisor: Prof. Dr. Felix Weidinger

  • Routenoptimierung eines Chinese postman problem am Beispiel der Darmstädter Schneeräumung

    2022

    Bachelor thesis

    Supervisor: Prof. Dr. Felix Weidinger

  • Column Generation für das Order Batching und Picking Problem

    2022

    Master thesis

    In Lagerhäusern ist die Reduktion von unproduktiven Wegzeiten eines des wichtigsten operativen Planungsproblemen. Das Zusammenfassen von Bestellungen zu sogenannten Batches, welche dann während derselben Tour kommissioniert werden, ist hierbei ein wichtiges Werkzeug, um Wegzeiten zu reduzieren. Ziel dieser Arbeit ist es ein Column Generation Verfahren zu entwickeln, welches das Order Batching Problem für ein gegebenes Lager Layout löst.

    Supervisor: Prof. Dr. Felix Weidinger

  • Umsatzoptimierung in Carsharing-Systemen durch Reinforcement Learning/Revenue Optimization in Car Sharing Using Reinforcement Learning

    2022

    Master thesis

    Supervisors: Prof. Dr. Felix Weidinger, Prof. Dr. Peter Buxmann

  • Ein Reinforcement Learning Ansatz für das Order Batching Problem in roboterbasierten Lagerhäusern

    2022

    Master thesis

    Supervisor: Prof. Dr. Felix Weidinger

  • Optimierung der Lagerplatzvergabe in Kommissionierlagern des Einzelhandels zur Minimierung der Laufwege

    2022

    Master thesis

    Supervisor: Prof. Dr. Felix Weidinger

  • Approach reinforcement learning to optimize batching for Order Batching Problem (OBP) in Manual Order Picking Systems

    2022

    Master thesis

    Supervisor: Prof. Dr. Felix Weidinger

  • Optimierung von Zuschnitt- und Falzprozessen mithilfe von mathematischer Modellierung

    2022

    Studienarbeit

    Supervisor: Prof. Dr. Felix Weidinger

  • Literaturüberblick zu Problemstellungen des stationsbasierten Car- und Bike-Sharing

    2022

    Bachelor thesis

    Ziel der Sharing-Economy ist es Ressourcen gemeinsam und damit effizienter zu nutzen. Besonders im Mobilitätssektor sind in den vergangenen Jahren viele neue Konzepte entstanden, welche durch wissenschaftliche Arbeiten begleitet wurden. Ziel dieser Arbeit ist es die Literatur zu Konzepten des Car- und Bike-Sharing zu sichten und zu systematisieren.

    Supervisor: Prof. Dr. Felix Weidinger

  • Entwurf und Implementierung eines Deep Reinforcement Learning Ansatzes als Eröffnungsheuristik für das Picker Routing Problem

    2021

    Master thesis

    Ziel der Arbeit ist es einen Deep Reinforcement Learning Ansatz zur Erzeugung einer ersten gültigen Lösung für das Picker Routing Problem zu entwerfen, zu implementieren und zu trainieren. Dieser soll anschließend mit traditionellen Verfahren in einer empirisch angelegten Studie verglichen werden.

    Supervisor: Prof. Dr. Felix Weidinger

  • Literaturüberblick zu Verfahren und Anwendungsszenarien des Deep Reinforcement Learning

    2021

    Bachelor thesis

    Das Deep Reinforcement Learning ist in aller Munde. Aber was ist es eigentlich? Ziel dieser Bachelorarbeit ist es das Thema zu erarbeiten und verschiedene Ansätze des Deep Reinforcement Learning zu verstehen und zu vergleichen.

    Supervisor: Prof. Dr. Felix Weidinger