"Since I'm personally very interested in research, it is really enjoyable for me to do a Ph.D. Not only can you study the topic you find fascinating, but you also get a decent scholarship and receive a highly valued degree after concluding your studies. I chose in particular to study at the computer science department in Paderborn due to matching research interests, as well as the department's and my supervisors' excellent reputation."– Peter Janacik, PACE PhD-student
Current student projects
Below you will find examples of short abstracts of the type of doctoral research currently being done by students at PACE. We hope this will give you an idea of the type of project you could be doing if you choose to complete your PhD in Paderborn.
Developments in recent years in the area of air cargo industry are characterized by continuous growth, which will hardly diminish in the future. This fact presents the operators of air cargo terminals with particular challenges. The cargo must be handled as efficient as possible, i.e. with a minimum of time and cost, in order to ensure the competitiveness of the companies. A crucial step in recent years to address these challenges represents the consequent automation of the terminals.
In light of the increasing automation, it seems surprising that the job scheduling is still being performed manually, though it is a key factor for the efficiency of the entire terminal. Up-to-date methods, e.g. from the field of operation research, are hardly used, because of frequently used restrictive assumptions, which apply in this form only rarely in practice. However, the manually created plans are often neither optimal nor robust in terms of the objective function. In the event of disruptions reactive adjustments to the plan are made that can resolve the problem in the short term, but due to the neglected plan stability may interfere with downstream processes. Due to the multitude of constraints which have to be considered and the complex optimization problem for even restrictive assumptions the manual creation of these plans is not effective.
In cooperation with an industrial partner, a world leader for air cargo logistic solutions, a novel approach for scheduling is developed, which transforms the hitherto manual scheduling process into an automated process. By using a proactive multi-objective optimization, which optimizes the objectives of timeliness, robustness and stability while maintaining orders, personnel, material handling and machine restrictions, a robust schedule is generated, which already takes into account known disruptions and immunizes the plan execution against them. In the case of qualitative/quantitative faults, which occurred for the first time, an adaptation of the currently running schedule in accordance with the plan stability is executed if necessary. Additionally, the deployment of personnel and the provision of the material at the workstations are included into the planning process. The design result is a schedule for the workstations and the staff, which can also be followed under the conditions of the stochastic system.
Storing goods is an essential time-transformation
process in production environments and overall supply chains. Warehouses
as central storage points permit to intercept uncertainty in supply and
Layout and dimensioning of warehouses are carried out most often by expert planners based on experience. Especially the dimensioning is often based on throughput estimations at bottlenecks like input/output points. This approach may lead to oversized and over-equipped solutions. Many decisions like the choice of loading devices are made without consideration of the influence to the whole system.
In research there are two main tracks regarding better warehouse-layout decisions: On the one hand there are approaches describing warehouse layout strategies but they are telling primary what to do in which sequence and not how. On the other hand many researchers deal with highly detailed problems like optimizing routing strategies for automated storage devices – but those often aren’t coupled to design decisions, or use assumptions that reduce the practicability quite strongly (e.g. only homogenous storage parts).
Goal of this research is the development of a decision support system, covering the whole design process for fully automated warehouses. This process consists of three main steps: First the determination of the amount of storage locations. Second the choice of loading devices and third the sizing of storage areas observing the influence of storage- and retrieval- performance with different systems and strategies. Finally an expense budgeting has to be integrated in order that the decision is not based solely on performance data.
Assembly lines are flow-oriented production systems, which gained an important practical role. One of the most important optimization problems for assembly line control is the sequencing problem.
Due to the varying installation complexity of options
the processing times of orders could be higher, equal or less than the capacity
of a workplace. For example the installation of a complex multimedia
system takes more time than the installation of a simple radio. Processing times higher than the capacity of the station require that
worker can work into surrounding stations – this is known as ‘drift’. If many
models with high processing times are processed one after the other, then the
worker must drift far away from his station border. At the same time the worker
positions are restricted to so called “drift limits”. If worker can’t finish
the processing of models within the drift limit, then compensation, e.g. line
stoppage or applying of cross trained utility workers, is necessary. These
overload situations are important cost driver and must be reduced to a minimum.
The impacts from decisions, made during the product life cycle of complex projects, especially in the air cargo terminal domain, are hard to estimate in the earlier planning phases. Such projects have strict restrictions given by several entities, like the customer or the government, that have to be fulfilled. Simulation guided decision making helps to find a reasonable solution for the given restrictions and the final material flow within an air cargo terminal.
The current status is, that for such specialized projects an extern company has to be hired to create the simulation model based on the experience of the planner. Therefor a lot of time and money will be consumed by the arrangement of the terms for the definitions of the machines and processes, developed for one particular air cargo terminal project.
In cooperation with a company this thesis focuses on the development of a simulation software specially designed for creating and measuring the performance of air cargo terminals. Such performance is depending on the static definitions, like physical restrictions, and the dynamical rearrangement of the parts which have an impact on the throughput and their process strategy.
The performance itself is calculated by the arrangement of the machines and the process strategy. And gives assistance for the selection of different planning characteristics for each machine and process but also for the whole system itself.
The global demographic change leads to an increasing number of elder people and therefore people in need of care. One possibility to support those people is home health care. The clients of home health care providers stay at their homes and require different services on different days mostly for a long period. As a result, the providers face complex routing and scheduling problems for planning the daily services.
The main contribution of this research project is to develop solution approaches for optimizing the routes and working schedules of the nurses. The current state of the art is extended with missing requirements from a practical point of view. These are e.g. working regulations and the consideration of different optimization goals at the same time. The focus lies on the customer and employee satisfaction. These are important for the providers to be competitive.
Another major point is the possibility to deal with dynamic sets of clients and employees. The need of rescheduling at certain points in time is crucial to practical application of the optimization approaches. The problem statement is solved with different exact and heuristic methods from the field of Operations Research. Especially the development of heuristic methods seems to be a promising solution, which can cope with complex real world data sets and deliver good schedules in reasonable time.
Today’s automotive companies need to focus much more on economic aspects then a couple of years ago. In order to reach this goal, the companies put a research focus on lightweight construction to reduce the total weight and thereby the fuel consumption of a car.
These lightweight material are challenging because the material structure is much more complex than it was before. The material engineers for carbon need to consider much more aspects during their work like the characteristics of the matrix system which surrounds the different fiber types and the structure, length and kind of fibers.
In order to analyze the material characteristics and to discover new facts, a data base with research and testing results, a data mining concept and a tool is needed to support the research work. In cooperation with the Daimler AG my PHD Project was created to develop a system which assumes this problem.