Filters

1 1
1 1 1
260
Reset filters

Settings

Publications Search Results

Now showing 1 - 10 of 260
  • Publication
    Reinforcement Learning for Two-Stage Permutation Flow Shop Scheduling - A Real-World Application in Household Appliance Production
    ( 2024)
    Müller, Arthur
    ;
    Grumbach, Felix
    ;
    Kattenstroth, Fiona
    Solving production scheduling problems is a difficult and indispensable task for manufacturers with a push-oriented planning approach. In this study, we tackle a novel production scheduling problem from a household appliance production at the company Miele & Cie. KG, namely a two-stage permutation flow shop scheduling problem (PFSSP) with a finite buffer and sequence-dependent setup efforts. The objective is to minimize idle times and setup efforts in lexicographic order. In extensive and realistic data, the identification of exact solutions is not possible due to the combinatorial complexity. Therefore, we developed a reinforcement learning (RL) approach based on the Proximal Policy Optimization (PPO) algorithm that integrates domain knowledge through reward shaping, action masking, and curriculum learning to solve this PFSSP. Benchmarking of our approach with a state-of-the-art genetic algorithm (GA) showed significant superiority. Our work thus provides a successful example of the applicability of RL in real-world production planning, demonstrating not only its practical utility but also showing the technical and methodological integration of the agent with a discrete event simulation (DES). We also conducted experiments to investigate the impact of individual algorithmic elements and a hyperparameter of the reward function on the overall solution.
  • Publication
    Quantencomputing in der maritimen Logistik: Vorstellung des MIRP Demonstrators
    ( 2023-11-24)
    Szal, Oliver
    Das Maritime Inventory Routing Problem (MIRP) ist ein komplexes mathematisches Optimierungsproblem, welches sich mit der effizienten Routenplanung von Schiffen zur Belieferung von verschiedenen Standorten befasst. Der MIRP Demonstrator ist eine von uns entwickelte Software, mit der man Probleminstanzen generieren und sowohl klassisch, als auch mit einem Quantenannealer lösen kann. Somit wird der Vergleich von klassischen und Quanten-Lösungen dieses Problems mit nur wenigen Maus-Klicks ermöglicht.
  • Publication
    A Reactive-Periodic Hybrid Optimization for Internal Hospital Logistics
    ( 2023-10-06)
    Ehsanfar, Ebrahim
    ;
    Karami, Farzaneh
    ;
    Kerkenhoff, Tim
    Internal hospital logistics (IHL) involves the scheduling of materials and patient transportations employing a fleet of transporters. The problem of collecting and delivering these items within a hospital can be modeled as a Pickup and Delivery Problem with Time Windows (PDPTW). This paper proposes a hybrid dynamic optimization to address the IHL problem based on a two-step heuristic. This algorithm combines reactive and periodic optimizations to assign logistic’s transports to the most suitable transporters while considering the urgency of each transport. To conserve resources, this algorithm addresses logistics transports with higher urgency reactively and handles less urgent transports periodically. The initial assignment is constructed using the earliest due date first (EDDF) assignment method. To further improve the efficiency of the procedure, a ruin and recreate heuristic is developed and tested. Computational experiments have been conducted utilizing hospital data from a large hospital with approximately 2100 beds located in Germany to evaluate the performance of the proposed dynamic hybrid optimization. Results show that the hybrid policy outperforms the baseline reactive policy used in the hospital in terms of service quality and cost efficiency.
  • Publication
    Location Problems with Cutoff
    ( 2023)
    Müller, Raoul
    ;
    Schöbel, Anita
    ;
    Schuhmacher, Dominic
    In this paper, we study a generalized version of the Weber problem of finding a point that minimizes the sum of its distances to a finite number of given points. In our setting, these distances may be cut off at a given value C > 0, and we allow for the option of an empty solution at a fixed cost C′. We analyze under which circumstances these problems can be reduced to the simpler Weber problem, and also when we definitely have to solve the more complex problem with cutoff. We furthermore present adaptions of the algorithm of Drezner, Mehrez and Wesolowsky (1991 [The facility location problem with limited distances. Transportation Science, 25(3), 183-187, INFORMS]) to our setting, which in certain situations are able to substantially reduce computation times as demonstrated in a simulation study. The sensitivity with respect to the cutoff value is also studied, which allows us to provide an algorithm that efficiently solves the problem simultaneously for all C > 0.
  • Publication
    Iterative Next Boundary Detection for Instance Segmentation of Tree Rings in Microscopy Images of Shrub Cross Sections
    ( 2023)
    Gillert, Alexander
    ;
    Resente, Giulia
    ;
    Anadon-Rosell, Alba
    ;
    Wilmking, Martin
    ;
    Lukas, Uwe Freiherr von
    We address the problem of detecting tree rings in microscopy images of shrub cross sections. This can be regarded as a special case of the instance segmentation task with several unique challenges such as the concentric circular ring shape of the objects and high precision requirements that result in inadequate performance of existing methods. We propose a new iterative method which we term Iterative Next Boundary Detection (INBD). It intuitively models the natural growth direction, starting from the center of the shrub cross section and detecting the next ring boundary in each iteration step. In our experiments, INBD shows superior performance to generic instance segmentation methods and is the only one with a built-in notion of chronological order. Our dataset and source code are available at http://github.com/alexander-g/INBD.
  • Publication
    Heuristics for a cash-collection routing problem with a cluster-first route-second approach
    ( 2023)
    Singh, Bismark
    ;
    Oberfichtner, Lena
    ;
    Ivliev, Sergey V.
    Motivated by a routing problem faced by banks to enhance their encashment services in the city of Perm, Russia, we solve versions of the traveling salesman problem (TSP) with clustering. To minimize the risk of theft, suppliers seek to operate multiple vehicles and determine an efficient routing; and, a single vehicle serves a set of locations that forms a cluster. This need to form independent clusters—served by distinct vehicles—allows the use of the so-called cluster-first route-second approach. We are especially interested in the use of heuristics that are easily implementable and understandable by practitioners and require only the use of open-source solvers. To this end, we provide a short survey of 13 such heuristics for solving the TSP, five for clustering the set of locations, and three to determine an optimal number of clusters—all using data from Perm. To demonstrate the practicality and efficiency of the heuristics, we further compare our heuristic solutions against the optimal tours. We then provide statistical guarantees on the quality of our solution. All of our anonymized code is publicly available allowing extensions by practitioners, and serves as a decision-analytic framework for both clustering data and solving a TSP.
  • Publication
    Simulation-Based Analysis of (Reverse) Supply Chains in Circular Product-Service-Systems
    ( 2023)
    Große Erdmann, Julian
    ;
    Koller, Jan
    ;
    Amir, Saman
    ;
    Mihelič, Aleš
    ;
    Döpper, Frank
    With an expected growth of global waste to 3.40 billion tonnes by 2050 and a circularity today of only 8.6% of the world, the earth’s sustainable resources are being exploited beyond their regeneration capacity. Hence, it is necessary to step away from a take - make - dispose principal and transform from a linear towards a circular economy to close product cycles to optimize resource consumption and reduce waste. Product-Service-Systems (PSSs), based on multiple product life cycles combined with remanufacturing, offer a solution to close product cycles. In such PSS, the responsibility for returning, remanufacturing, and repairing used products remains with the Original Equipment Manufacturer (OEM) and increases its need in (reverse) supply chain activities. Essential factors for (reverse) supply chains are, e.g., determining the distribution network, the location of recovery facilities, the geographical dispersion of the customers, and the information flows between the different stakeholders. In this context, this work proposes a multi-method simulation model to support practitioners in determining the optimal infrastructure for storing, remanufacturing, and repairing the used products regarding economic and ecological target criteria. The applicability of the proposed approach is illustrated through a case study of a white goods manufacturing company. This case study highlights the importance of determining the optimal infrastructure in a (reverse) supply chain in PSS business models.
  • Publication
    EETTlib - Energy-efficient train timetabling library
    ( 2023)
    Bärmann, A.
    ;
    Gemander, Patrick
    ;
    Hager, L.
    ;
    Nöth, F.
    ;
    Schneider, Oskar
    We introduce EETTlib, an instance library for the Energy-Efficient Train Timetabling problem. The task in this problem is to adjust a given timetable draft such that the energy consumption of the resulting railway traffic is minimized. To this end, the departure times of the trains can be slightly, and their velocity profiles on each trip can be modified. We provide real-world data originating from two research projects in this field, one with Deutsche Bahn AG, the most important railway company in Germany, the other with VAG Verkehrs-Aktiengesellschaft, the operator of public transport in the city of Nürnberg, Germany. In both cases, our library contains representative data on the relevant operational constraints and supports various possible choices for the objective function with respect to energy-efficiency. The resulting benchmark instances can be used by the scheduling and timetabling community to improve their models and algorithms. They are available under https://www.eettlib.fau.de.
  • Publication
    Asymptotic behavior for textiles with loose contact
    ( 2023)
    Orlik, Julia
    ;
    Falconi, Riccardo
    ;
    Griso, Georges
    ;
    Wackerle, Stephan
    The paper is dedicated to the modeling of the elasticity problem for a textile structure. The textile is made of long and thin fibers, crossing each other in a periodic pattern, forming a woven canvas of a square domain. The textile is partially clamped. The fibers cannot penetrate each other but can slide with respect to each other in the in-plane directions. The sliding is bounded by a contact function, which is chosen loose. The partial clamp and the loose contact lead to a domain partitioning, with different expected behaviors on each of the four subdomains. The homogenization is made via the periodic unfolding method, with an additional dimension reduction. The macroscopic limit problem results in a Leray-Lions problem with only macroconstraints in the plane.
  • Publication
    BatNet: a deep learning-based tool for automated bat species identification from camera trap images
    ( 2023)
    Krivek, Gabriella
    ;
    Gillert, Alexander
    ;
    Harder, Martin
    ;
    Fritze, Marcus
    ;
    Frankowski, Karina
    ;
    Timm, Luisa
    ;
    Meyer‐Olbersleben, Liska
    ;
    Lukas, Uwe Freiherr von
    ;
    Kerth, Gerald
    ;
    Schaik, Jaap van
    Automated monitoring technologies can increase the efficiency of ecological data collection and support data-driven conservation. Camera traps coupled with infrared light barriers can be used to monitor temperate-zone bat assemblages at underground hibernacula, where thousands of individuals of multiple species can aggregate in winter. However, the broad-scale adoption of such photo-monitoring techniques is limited by the time-consuming bottleneck of manual image processing. Here, we present BatNet, an open-source, deep learning-based tool for automated identification of 13 European bat species from camera trap images. BatNet includes a user-friendly graphical interface, where it can be retrained to identify new bat species or to create site-specific models to improve detection accuracy at new sites. Model accuracy was evaluated on images from both trained and untrained sites, and in an ecological context, where community- and species-level metrics (species diversity, relative abundance, and species-level activity patterns) were compared between human experts and BatNet. At trained sites, model performance was high across all species (F1-score: 0.98-1). At untrained sites, overall classification accuracy remained high (96.7-98.2%), when camera placement was comparable to the training images (<3 m from the entrance; <45° angle relative to the opening). For atypical camera placements (>3 m or >45° angle), retraining the detector model with 500 site-specific annotations achieved an accuracy of over 95% at all sites. In the ecological case study, all investigated metrics were nearly identical between human experts and BatNet. Finally, we exemplify the ability to retrain BatNet to identify a new bat species, achieving an F1-score of 0.99 while maintaining high classification accuracy for all original species. BatNet can be implemented directly to scale up the deployment of camera traps in Europe and enhance bat population monitoring. Moreover, the pretrained model can serve as a baseline for transfer learning to automatize the image-based identification of bat species worldwide.