Current Bachelor & Master Thesis Topics

The Production & Logistics Networks Workgroup is currently expressly not offering any more thesis topics except the ones listed below. If you are interested, please read the hints on how to write a thesis with the PLN workgroup and contact the supervisor mentioned below.

Bachelor or Master Topics

Triggers and Effects of Synchronization in Supply Chains – Minimal Model Investigation
Synchronization in a broader sense means aligning a certain behavior or state over time. In the physical world, two systems can spontaneously synchronize with respect to each other even if there are only small physical interactions. Up to now it is not fully understood how synchronization affects supply chains: Does a spontaneous synchronization also occur in supply chains? If so, what triggers it and how does it influence the achievement of the logistics objectives and the robustness of the processes?
The purpose of this thesis is to study synchronization phenomena present in supply chains. Minimal models can be used to investigate the triggers and effects of those phenomena occurring in supply chains.  By applying synchronization measures to the models, the thesis will firstly examine the relation between synchronization and the supply chain’s core characteristics (e.g., design, structure, operations and control strategy, arrival times (distribution), processing time (distribution)). Secondly, it will research the link between synchronization and the supply chain performance (e.g., lateness, inventory levels) and robustness (e.g., deviations in performance due to system disturbances), thus drawing conclusions for the consequences that synchronization has on supply chains. A good starting point is:

  • View document on publisher site S. M. Chankov, M. Hütt, and J. Bendul, “Synchronization in manufacturing systems: quantification and relation to logistics performance,” International Journal of Production Research, 2016.
    [Bibtex]
    @article{Chankov2016a,
    author = {Chankov, Stanislav M. and H{\"u}tt, Marc-Thorsten and Bendul, Julia},
    title = {Synchronization in manufacturing systems: quantification and relation to logistics performance},
    journal = {International Journal of Production Research},
    doi = {10.1080/00207543.2016.1165876},
    URL = {http://dx.doi.org/10.1080/00207543.2016.1165876},
    eprint = {http://dx.doi.org/10.1080/00207543.2016.1165876},
    year={2016},
    abstract = {The term 'synchronization' in manufacturing refers to the provision of the right components to the subsequent production steps at the right moment in time. It is widely assumed that synchronization is beneficial to the logistics performance of manufacturing systems. However, it has been shown that synchronization phenomena can be detrimental to systems in which they emerge. To study if synchronization phenomena also occur in and affect manufacturing systems' performance, a formal quantification and holistic understanding of the types of synchronization phenomena emerging in manufacturing are needed. This article aims to fill this research gap by developing synchronization measures for manufacturing systems, applying these measures to real-world production feedback data and utilising them to test the assumption about synchronization's beneficial effect on logistics performance. We identify two distinct synchronization types occurring in manufacturing systems, logistics and physics synchronization, and show that they are negatively correlated. Further, we show that logistics synchronization and due date performance exhibit anti-correlation and thus question the assumption that synchronization leads to higher efficiency in manufacturing systems. This article aids production managers in designing and optimising production systems, and supports further empirical research in production planning and control and production system design.},
    }
  • View document on publisher site S. M. Chankov, G. Malloy, and J. Bendul, “The Influence of Manufacturing System Characteristics on the Emergence of Logistics Synchronization: A Simulation Study,” in Dynamics in Logistics: Proceedings of the 5th International Conference LDIC, 2016 Bremen, Germany, M. Freitag, H. Kotzab, and J. Pannek, Eds., Cham: Springer International Publishing, 2016, pp. 29-40.
    [Bibtex]
    @inbook{Chankov2016b,
    author="Chankov, Stanislav M. and Malloy, Giovanni and Bendul, Julia",
    editor="Freitag, Michael and Kotzab, Herbert and Pannek, J{\"u}rgen",
    title="The Influence of Manufacturing System Characteristics on the Emergence of Logistics Synchronization: A Simulation Study",
    bookTitle="Dynamics in Logistics: Proceedings of the 5th International Conference LDIC, 2016 Bremen, Germany",
    year="2016",
    publisher="Springer International Publishing",
    address="Cham",
    pages="29--40",
    isbn="978-3-319-45117-6",
    doi="10.1007/978-3-319-45117-6_3",
    url="http://dx.doi.org/10.1007/978-3-319-45117-6_3"
    abstract="The term 'synchronization' in manufacturing refers to the provision of the right components to the subsequent production steps at the right moment in time. It is still unclear how manufacturing system characteristics impact synchronization. Thus, the purpose of this paper is to investigate the effect of manufacturing systems' characteristics on the emergence of logistics synchronization in them. We conduct a discrete-event simulation study to examine the effect of three system characteristics: (1) material flow network architecture, (2) work content variation, and (3) order arrival pattern. Our findings suggest that the material flow network architecture and the work content variation are related to logistics synchronization. Linear manufacturing systems with stable processing times such as flow shops operate at high logistics synchronization levels, while highly connected systems with high variability of processing times such as job shops exhibit lower synchronization levels."
    }
  • View document on publisher site M. A. Schipper, S. M. Chankov, and J. Bendul, “Synchronization Emergence and its Effect on Performance in Queueing Systems,” Procedia CIRP, vol. 52, pp. 90-95, 2016, The Sixth International Conference on Changeable, Agile, Reconfigurable and Virtual Production (CARV2016).
    [Bibtex]
    @article{Schipper:2016,
    title = "Synchronization Emergence and its Effect on Performance in Queueing Systems ",
    journal = "Procedia CIRP",
    volume = "52",
    number = "",
    pages = "90 -- 95",
    year = "2016",
    note = "The Sixth International Conference on Changeable, Agile, Reconfigurable and Virtual Production (CARV2016) ",
    issn = "2212-8271",
    doi = "10.1016/j.procir.2016.07.016",
    url = "http://www.sciencedirect.com/science/article/pii/S2212827116307582",
    author = "Schipper, Manuel A. and Chankov, Stanislav M. and Bendul, Julia",
    keywords = "synchronization",
    keywords = "queueing theory",
    keywords = "production system",
    keywords = "manufacturing system design ",
    abstract = "Abstract Synchronization as a dynamic process has found applications in many fields. However, it remains unclear how this phenomenon relates to manufacturing systems. The aim of this study is to investigate the conditions for emergence of synchronization and its effects on the wide spectrum of production logistics performance objectives. Using queueing theory as the underlying methodology for deductive modeling of manufacturing systems, we run computer simulations on networks of queueing systems and investigate synchronization measurements in relation to system parameters and performance indicators. Our initial findings suggest that different types of manufacturing systems display different synchronization behaviors and that periodically driven systems with deterministic arrival and service rates display higher synchronization in comparison to stochastic ones. Further, we show that intrinsic physics synchronization is correlated to capacity utilization, throughput times and WIP levels, suggesting the co-activity of operations is related to highly utilized systems, while external physics synchronization is anticorrelated to throughput times and WIP levels, suggesting that higher efficiencies emerge with workstation repetitive behavior."
    }
 Stanislav Chankov
What is Really “On-Time”? A Comparison of Due Date Performance Indicators
One of the most important logistics objective of companies is schedule reliability, i.e. meeting the due dates set by customers. On-time delivery is essential in today’s dynamic conditions: if a company cannot produce and deliver on time, it has to make up for it by incurring a penalty on delays and using high cost express delivery. If that option is not available, customer dissatisfaction is inevitable. When the customer’s production system operates with low inventories and just-in-time deliveries, low schedule reliability of the suppliers will eventually ruin that customer’s schedule reliability as well. In spite of this high importance, industry analyses show that the schedule reliability of many factories is quite low: a part of the orders is completed late and endangers the delivery reliability of the company; other orders are completed before the scheduled date and increase finished goods inventories (Lödding et al. 2014). The question of how to measure schedule reliability arises.

The purpose of this thesis is to compare several measures for due date performance and study their interrelationships. Using an existing simulation model, different experiments are run to generate data. Statistical analyses (e.g. Pearson’s Correlation Analysis, t-test, ANOVA) on the generated data sets are performed as part of the course of research to compare different due date performance measures. A good starting point is

  • View document on publisher site R. Schäfer, S. Chankov, and J. Bendul, “What is Really “On-Time”? A Comparison of Due Date Performance Indicators in Production,” Procedia CIRP, vol. 52, pp. 124-129, 2016, The Sixth International Conference on Changeable, Agile, Reconfigurable and Virtual Production (CARV2016).
    [Bibtex]
    @article{Schaefer:2016,
    title = "What is Really ``On-Time''? A Comparison of Due Date Performance Indicators in Production ",
    journal = "Procedia CIRP",
    volume = "52",
    number = "",
    pages = "124 -- 129",
    year = "2016",
    note = "The Sixth International Conference on Changeable, Agile, Reconfigurable and Virtual Production (CARV2016) ",
    issn = "2212-8271",
    doi = "10.1016/j.procir.2016.07.017",
    url = "http://www.sciencedirect.com/science/article/pii/S2212827116307594",
    author = "Ricarda Sch{\"a}fer and Stanislav Chankov and Julia Bendul",
    keywords = "due date performance",
    keywords = "lateness",
    keywords = "production planning and control ",
    abstract = "Abstract On-time delivery is essential in today's dynamic conditions: if a company cannot produce and deliver on time, it has to make up for it by using high cost express delivery or faces customer dissatisfaction. One factor influencing the delivery reliability is the due date performance (DDP) within production. Although the significance of DDP has been established, the question of how to measure it remains. A review of existing literature shows the vast amount of different DDP measures (lateness, relative lateness, tardiness, schedule reliability, etc.). The purpose of this paper is to compare different DDP measures used in manufacturing in order to assess their interrelationship, so that companies are better able to understand the impact of their choice of measure. A review of DDP measures described in literature is performed, followed by statistical analysis of the relations between those measures computed on production feedback data from four real-world manufacturers. The results indicate that there exist differences across DDP measure groups. Further research is needed to assess the benefits of each measure in a given situation."
    }

.

 Stanislav Chankov
Crowd Logistics: Crowdsourcing Potential for the Logistics Industry
Crowdsourcing is an increasingly used concept in a variety of fields, mainly within the scope of the sharing/collaborative economy. Crowdsourcing practices have been successfully used in other key economic fields such as: professional/personal services (e.g. Craigslist), pre-owned goods (e.g. eBay), custom products (e.g. Etsy), funding (e.g. Kickstarter), money-lending (e.g. Kiva) or transportation (e.g. Uber). While companies like Uber are already well-established in the passenger transportation area, crowd logistics (e.g. crowdsourced goods delivery, crowdshipping, or shared storage) has only recently received interest in both theory and practice. Amazon has started its initiative in Seattle September 2015 through Amazon Flex and Uber is already operating a delivery service since April 2015 named UberRUSH.

Crowdsourcing could potentially be very suitable for logistics services. For example, it can help for last-mile shipments, which are becoming more and more problematic mostly due to the high growth of e-commerce. The aim of this study is to investigate crowdsourcing applications for logistics services. The course of research can involve conducing an online survey to collect data or extensive online search to identify business models for crowd logistics. A good starting point is Rougès and Montreuil (2014).

A topic for the general sharing/collaborative economy (e.g. Airbnb) is also possible, you can come up with your own ideas.
 Stanislav Chankov
Improving logistics reporting through visualization and avoidance of cognitive biases
This is a thesis project in cooperation with ArcelorMittal Bremen as part of a paid, seven month (approximately), paid working student position at the company. The thesis project aims at improving the reports generated by a planning for different audiences (functions in the company).
The project is supervised by our former colleague Mathias Knollmann and applications have to be submitted to ArcelorMittal.
More information (in German) can be found here.
Dr. Mathias Knollmann (please use contact data from PDF file)

For further information on the bachelor thesis seminar, please see the website of the Industrial Engineering and Management (BSc) program.

Bachelor or Master Topics supervised by Prof. Hütt

Prof. Hütt (Professor of Systems Biology at Jacobs University) has agreed to supervise the following thesis topics. Please approach him directly in case of any questions.

Time constants of biological enzymes as machine capacities
Metabolism is a production process in biological cells, where enzymes (the ‘machines’) convert chemical compounds (metabolites) into other chemical compounds, leading to an intricate network of interacting machines. The striking parallel between metabolism and industrial production systems offers the fascinating possibility to compare evolved and designed systems. As ever more information on metabolic systems becomes electronically available in bioinformatics databases, this parallel can now be quantitatively explored. The goal of the project is to use production planning algorithms for machine capacities, apply them to a metabolic network and see, whether the actual capacities (given by the inverse time constants of enzymes and available via databases) match those predictions.

Further reading:
Beber, M. E. and Hütt, M. (2012). How do production systems in biological cells maintain their function in changing environments? Logistics Research, 5(3-4):79–87.

Prof. Hütt
Product diversity and network structure: a minimal model

A comparatively unexplored aspect of manufacturing is the relationship between the production network and the diversity of the manufactured products. Previous work on evolved flow networks (Beber et al. 2013) has shown that the network architecture is strongly affected by the complexity of the network’s required ‘output pattern’. In a series of small investigation we want to understand this phenomenon more deeply and apply the findings to manufacturing.
The goal of the project component described here is to formulate heuristics for generating a production network for a given product structure using path combinatorics. Then the network architectures will be statistically analyzed in order to establish relationships between product diversity and network structure.

Further reading:
Beber, M., Armbruster, D. and Hütt, M.-Th. (2013) The prescribed output pattern regulates the modular structure of flow networks. European Physical Journal B 86, 473. DOI: 10.1140/epjb/e2013-40672-3.

Prof. Hütt
Product diversity and network structure: analysis of evolved flow networks

A comparatively unexplored aspect of manufacturing is the relationship between the production network and the diversity of the manufactured products. Previous work on evolved flow networks (Beber et al. 2013) has shown that the network architecture is strongly affected by the complexity of the network’s required ‘output pattern’. In a series of small investigation we want to understand this phenomenon more deeply and apply the findings to manufacturing.
The goal of the project component described here is to analyze the database from Beber et al. (2013) from the perspective of novel network properties: linearity, path combinatorics and betweenness centrality.

Further reading:
Beber, M., Armbruster, D. and Hütt, M.-Th. (2013) The prescribed output pattern regulates the modular structure of flow networks. European Physical Journal B 86, 473.

Prof. Hütt
Product diversity and network structure: analysis of real production networks

A comparatively unexplored aspect of manufacturing is the relationship between the production network and the diversity of the manufactured products. Previous work on evolved flow networks (Beber et al. 2013) has shown that the network architecture is strongly affected by the complexity of the network’s required ‘output pattern’. In a series of small investigation we want to understand this phenomenon more deeply and apply the findings to manufacturing.
The goal of the project component described here is to study, whether similar relationships are also observed in real manufacturing systems. To this end, product diversity will be analyzed and then statistically compared with the system’s material flow networks.

Further reading:
Beber, M., Armbruster, D. and Hütt, M.-Th. (2013) The prescribed output pattern regulates the modular structure of flow networks. European Physical Journal B 86, 473.

Prof. Hütt
Synchronization as a quantifier of activity patterns: delay avalanches

The various concepts of synchronization attempt to capture the pattern of activity in a production network. It has been established before that high synchronization can lead to system-wide failures and thus reduce the robustness of the system against perturbations (Fretter et al. 2010). This project uses the general model paradigm of ‘avalanches on graphs’ to study this phenomenon further: How does the network architecture facilitate such avalanches? Can we adapt standard avalanche models to account for lateness propagation in production systems?

Further reading:
Fretter, C., Krumov, L., Weihe, K., Müller-Hannemann, M. and Hütt, M.-Th. (2010) Phase synchronization in railway timetables, European Physical Journal B 77, 281-289.

Prof. Hütt
Synchronization as a quantifier of activity patterns: activity as excitable dynamics

The various concepts of synchronization attempt to capture the pattern of activity in a production network. It has been established before that high synchronization can lead to system-wide failures and thus reduce the robustness of the system against perturbations (Fretter et al. 2010). This project asks, how different types of synchronization (termed logistics synchronization and physics synchronization in Chankov et al. 2015) quantify activity patterns in a graph, and thus ‘calibrate’ these measures. In order to understand the generic properties of these synchronization measures, we will use a simple model of activity, namely excitable dynamics on graphs (see, e.g., Müller-Linow et al. 2008) to simulate activity patterns and then analyze the synchronization measures as a function of network architecture and parameters of the dynamics.

Further reading:
Fretter, C., Krumov, L., Weihe, K., Müller-Hannemann, M. and Hütt, M.-Th. (2010) Phase synchronization in railway timetables, European Physical Journal B 77, 281-289.
Chankov, S., Bendul, J. and Hütt, M.-Th. (2015) Synchronization in Manufacturing Systems: Quantification and Relation to Logistics Performance. International Journal of Production Research, under review.
Müller-Linow, M., Hilgetag, C. and Hütt, M.-Th. (2008) Organization of excitable dynamics in hierarchical biological networks. PLoS Computational Biology 4, e1000190.

Prof. Hütt
Sequentiality and linearity of production networks

Important properties of manufacturing processes are determined by the combinatorics of paths in the material flow network (see Garcia et al. 2014 for an example, how such a question is investigated for closed paths). Deterministic linear sequences in this material flow ‘decouple’ parts of the system from other parts. The amount of linearity in a production network is therefore an important architectural quantity. While there is no standard way of measuring the linearity of a network, a rich set of network quantifiers has emerged over the last years addressing aspects of this question. The purpose of this project is to perform a literature review of these quantifiers and then compare them using a small set of reference networks with varying amounts of linearity.

Further reading:
Garcia, G.C., Lesne, A., Hilgetag C.C. and Hütt, M.-Th. (2014) The role of topological cycles in excitable dynamics on graphs. Phys. Rev. E 90, 052805.

Prof. Hütt
The graph chromatic number as a robustness indicator of production networks

The graph coloring problem (distribute colors from a list on a graph such that no same colors are linked) is related to many scheduling problems in logistics. Attempting to distribute the colors based on local decisions only generates coloring dynamics, which are a minimal model of autonomous control (see, e.g., Windt and Hütt 2010). Given a graph, the minimal number of colors for which the graph coloring problem can be solved is called the ‘chromatic number’ of the graph. Qualitatively speaking, this quantity determines, how easy scheduling is on the graph. We can expect that logistics performance of, e.g., a production network will depend strongly on this chromatic number. Using numerical experiments with scheduling software this relationship will be explored.Further reading:
Windt, K. and Hütt, M.-Th. (2010) Graph Coloring Dynamics: A Simple Model Scenario for Distributed Decisions in Logistics. CIRP Annals Manufacturing Technology 59, 461-464.
Prof. Hütt
Network recovery: a literature review with small numerical experiments

How networks recover from perturbations is a general question with deep implications for logistics systems. As an example, in Hao et al. (2015) the performance of a distributed insurance system under spatially and temporally correlated failures has been studied. The goal of this project is to understand, how the recovery of (production or distribution) networks is influenced by their architecture. The starting point will be a literature review of attempts to quantify and analyze network recovery. This survey of theoretical studies will be complemented by own numerical experiments on network recovery.

Further reading:
Hao, Y., Armbruster, D. and Hütt, M.-Th. (2015) Node survival in networks under correlated attacks. PLoS One, in press.

Prof. Hütt
Network representations of production systems

Across many disciplines, the formal language of nodes and links provides an efficient data structure for representing complex systems. Such representations can help comparing diverse systems.  In the case of production systems, nodes can be machines, processes, inventories, products at intermediate stages of production, or check points. Links can represent material flow, regulation, control and decision alternatives.
As an example, in Becker et al. (2011) production systems are represented as networks of cyclically operating devices. In this way, a comparison with traffic networks and production systems in biological cells (metabolic networks) could be achieved.
The goal of this project is to review the different network representations of production systems and analyze, how these network representations can help understand the functioning of these systems.

Further reading:
Becker, T., Beber, M.E., Windt, K., Hütt, M.-Th. and Helbing, D. (2011) Flow control by periodic devices: A unifying language for the description of traffic, production, and metabolic systems. J. Stat. Mech, P05004.

Prof. Hütt
Production networks vs. information processing networks

Across many disciplines, the formal language of nodes and links provides an efficient data structure for representing complex systems. Such representations can help comparing diverse systems.
Such networks tend to fall into two classes: production networks, characterized by their material flow, and information processing networks, characterized by their flow of information.
The goal of this project is a high-level comparison of these two classes of networks: What are the main functional differences (e.g., conservation laws)? Do they differ systematically in their architectures? How do the functional requirements differ for these networks? What do efficiency and robustness mean in each class?

Prof. Hütt
Random walks as a reference model for material flows

Understanding the material flow in manufacturing systems and its impact on logistics performance indicators is one of the principal goals production logistics. In many complex systems, a proper analysis of available data is only possible when contrasted to suitable versions of random data. Here, a simple model for random material flows, based on random walks, will be explored. Over the last decade, random walks on graphs have dramatically enhanced our understanding of the scaling of fluctuations in networks (see, e.g., Kosmidis et al. 2015). The goal of this project is to introduce modifications to a standard random walk (like a preference to move from an input layer to an output layer of the network) and study, how well the random walk reproduces statistical features of real material flows in manufacturing.

Further reading:
Kosmidis, K., Beber, M. and Hütt, M.-Th. (2015) Network heterogeneity and node capacity lead to heterogeneous scaling of fluctuations in random walks on graphs. Advances in Complex Systems 18, 1550007.

Prof. Hütt
Scaling laws in production logistics

Scaling relationships are among the most surprising findings about complex social and technological systems: laws allowing to predict with high accuracy the number of patents, crimes and restaurants just from the population size (see Bettencourt and West 2010); laws relating the number of machines to the number of regulators in production systems in biological cells (Maslov et al. 2009). The goal of this project is the search for such scaling laws in production systems. The starting point will be a survey of publically available databases about company sizes and infrastructures.

Further reading:

Bettencourt, L, and Geoffrey West, G.B. (2010) A unified theory of urban living. Nature 467, 912-913.

Maslov, S., Krishna, S., Pang, T. and Sneppen, K. (2009) Toolbox model of evolution of prokaryotic metabolic networks and their regulation. PNAS 106, 9743.

 Prof. Hütt