Marvin Kastner

Marvin Kastner
Technische Universität Hamburg | TUHH · Institute of Maritime Logistics

M. Sc. Informatics

About

10
Publications
1,706
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
13
Citations
Introduction
Marvin Kastner currently works on creating a planning framework to automate the design of container terminals including the layout and equipment selection. The framework uses metaheuristics and the simulation-based optimization approach for discrete-event simulation models to create and evaluate competing designs. In addition, he is greatly interested in employing data science (e.g., explorative visualizations, machine learning, and statistics) for creating new insights into industrial data.

Publications

Publications (10)
Chapter
In maritime logistics, mathematical optimization and simulation are widely-used methods for solving planning problems and evaluating solutions. When putting these solutions to test, extensive and reliable data are urgently needed but constantly scarce. Since comprehensive real-life data are often not available or are classified as sensitive busines...
Conference Paper
Full-text available
Purpose: In the last decades, the transport capacities of container vessels have tremendously increased. This leads to longer berth times and greater peak loads at container terminals, especially when schedules are perturbed. Thus, existing container handling processes need to be re-evaluated regarding their adequacy. Methodology: In the first step...
Article
Full-text available
At container terminals, many cargo handling processes are interconnected and occur in parallel. Within short time windows, many operational decisions need to be made and should consider both time efficiency and equipment utilization. During operation, many sources of disturbance and, thus, uncertainty exist. For these reasons, perfectly coordinated...
Preprint
Full-text available
At container terminals, many cargo handling processes are interconnected and take place in parallel. Within short time windows, many operational decisions need to be taken considering both time and equipment efficiency. During operation, many sources for disturbance exist. These are the reason why perfectly coordinated processes are possibly unrave...
Conference Paper
Full-text available
Purpose: When container terminals are planned or converted, among others the most suitable container handling system needs to be selected and the appropriate terminal layout needs to be designed. These two planning activities are mutually de-pendent and affect the costs and future operational performance. This leads to the question of how to arrive...
Preprint
Full-text available
Purpose: When container terminals are planned or converted, among others the most suitable container handling system needs to be selected and the appropriate terminal layout needs to be designed. These two planning activities are mutually dependent and affect the costs and future operational performance. This leads to the question of how to arrive...
Conference Paper
Full-text available
Teaching machine learning in fields outside of computer sciences can be challenging when the students do not have a solid code knowledge. In this work, the requirements for teaching data literacy and code literacy to students of logistics are explored. Specifically, the use of Jupyter Notebooks in a machine learning course for students in logistics...
Chapter
Container terminals are highly complex systems where various processes need to interoperate smoothly in order to offer the required level of service at competitive prices. Hence, with changing logistics network structures, higher expected service standards, and increasing transportation demands, container terminals continuously need modification. W...
Conference Paper
Full-text available
Purpose: While simulation-based optimization has been discussed in theory and practically employed at container terminals, the different publications in this field have not yet been presented and compared in a structured manner. This paper gathers the latest developments and examine the similarities and differences of the provided approaches. Furth...
Poster
Full-text available
The learning outcome of the interdisciplinary master module „machine learning in logistics“ is the ability to visualize, clean, and interpreting big data, as well as identifying connections with methods of machine learning. The media-didactical challenge is to make machine learning accessible for those students who do not possess sound programming...

Questions

Questions (2)
Question
I am currently searching for information about the Ka-band satellite link quality. It should be machine-readable real-world data. I am looking at the link quality from the developer view so I am interested in the actual connection problems for a given common network configuration.
So what I'd like to have is a (preferably freely available) dataset of the re-transmitted packages (if TCP) or dropped packages (if UDP) at a given time and place (so I could combine it with some weather data). Round trip times could also be nice.
Is there such a dataset out there? Is it shared under a free license?
Question
The general setting is like this: There are lots of barometers, of some I have a verified position but others might have moved in the past. Those ones I want to discard.
My idea is, that if they moved for large distances, parts of their measurement series should be quite distinct from the observations of the verified barometers. But how can I determine that threshold?
My idea is that it could be done on the base of physics as there can't be too different pressure records at the same time spatially close to each other. Any suggestions?

Network

Cited By

Projects

Projects (4)
Archived project
Layout planning and logistics simulation have so far been used separately for the development and planning of logistics nodes. Layout planning software is characterized by an intuitive and cooperative application, but only allows static considerations. Simulation software, on the other hand, has the advantage of allowing the analysis of dynamic processes, but is less intuitive to use. The goal of ISI-Plan was to combine both tools and thus to use the strengths of the tools, to eliminate the weaknesses and to realize synergy effects (e.g. reduction of the modeling time for the simulation model).