Category Shine a light on ...
02 December 2022

Congratulations Florian Sachs! The first doctoral researcher successfully completed his doctorate at TUM Campus Heilbronn.

Tag Heilbronn
Tag Success Stories



At the TUM Heilbronn Campus, we congratulate our first doctoral researcher, Florian Sachs from the Center for Digital Transformation, on receiving his Ph.D.!


In his dissertation titled Integrated planning of unreliable flow lines with limited buffer capacities and spare parts provisioning, Florian Sachs analyzed unreliable production systems. He quantified the potential savings of an integrated planning approach in which spare parts inventory and production system setup are planned simultaneously – instead of sequentially, as in the past. He was the first to perform such a joint analysis and found that spare parts tend to be more effective when located at or near the center of the production system. In addition, he showed that the use of standardized components provides substantial cost savings.


“The TUM Campus Heilbronn, with our team's distinct expertise and internationality, provided me with an excellent environment for my Ph.D. In addition, the workshops at our campus offered me numerous opportunities to meet outstanding international researchers and discuss impactful research collaborations,” says Florian Sachs.

Prof. Gudrun Kiesmüller, Professor for Operations Management and Center Director at the Center for Digital Transformation, supervised Florian Sachs’ thesis. “Florian Sachs did an excellent doctoral thesis and, similar to our Center, has grown with the challenges. I wish him all the best for the next steps in his professional career”, she states.


A big thank you also goes to Prof. Geert-Jan van Houtum, Professor of Maintenance and Reliability at Technical University Eindhoven, who acted as a second reviewer and was part of the committee as well.


Florian Sachs will continue his scientific career and become an assistant professor for Supply Chain Analytics at the University of Cologne in spring 2023. Congratulations on this as well! We wish you all the very best and continued success!


Read more

Excerpts from Florian Sachs’ dissertation have already been published in separate papers:

  • F.E. Sachs, S. Helber, G.P. Kiesmüller. „Evaluation of Unreliable Flow Lines with Limited Buffer Capacities and Spare Part Provisioning”, European Journal of Operational Research, Volume 302, Issue 2, 2022, Pages 544-559.
  • G.P. Kiesmüller, F.E. Sachs. “Spare parts or buffer? How to design a transfer line with unreliable machines”, European Journal of Operational Research, Volume 284, Issue 1, 2020, Pages 121-134.
  • In this thesis, we consider different models of unreliable flow lines with intermediate buffers decoupling adjacent production stages. Each machine contains exactly one unit of a failure-prone critical component. In case of a component failure, we assume that the component units can be replaced by ready-to-use spare parts, thus achieving a high machine availability. For the stock-keeping of spare parts, we assume a one-for-one replenishment policy.

    We present a model of a two-machine flow line with stochastic processing times and develop a continuous-time Markov chain to exactly evaluate the essential system characteristics like average throughput or inventories. Based on that, we present a novel decomposition approach to approximately compute the flow-line characteristics for systems with stochastic processing times consisting of an arbitrary number of machines, limited buffer capacity, and spare parts. By means of extensive numerical examples, we analyze the effects of system parameters on approximation quality. We conclude that our method yields striking accuracy after comparing with results obtained by a Markov approach for smaller lines and discrete-event simulation for longer lines. Our results indicate the existence of complex interaction and partial substitution effects between buffer capacity and spare part base-stock levels.

    To answer the question of how to design a flow line with buffers and spares, we study the buffer and spare part allocation problem. Since the buffer allocation problem is NP-hard, we focus on heuristic solutions. We propose three greedy heuristics and apply the two metaheuristics, simulated annealing and a genetic algorithm. We compare all heuristics to complete enumeration for small flow lines in order to find the best-suited algorithm to apply to longer flow lines. We analyze different balanced and unbalanced flow line scenarios using a large-scale numerical study. First, we find that spare parts tend to be more effective when arranged at or near the center of a flow line already known for buffers. Second, we quantify the interaction between buffers and spare parts and amplify the strong cost dependence of the best flow line design. Third, we identify that results from the literature on spare parts planning still apply but that planners need our new approach for quantification because of the mentioned interactions.

    Using a model of a two-machine system, we study the impact of deterministic and fixed processing times. We calculate system characteristics with a discrete-time Markov chain and apply complete enumeration because the search space size allows for this optimization procedure. We observe that our results are still applicable and get even more considerable if the processing time variability is not present. Furthermore, we compare two system models: one with different critical components, as before, and one with identical components. By comparing both cases, we can identify the impact of component standardization regularly observed in practice. It turns out that component commonality can render significant cost savings possible. The reason for this is, on the one hand, the pooling effect of safety stocks and, on the other hand, the possibility of reducing buffer capacity.