Degree theses

Dept. VIII: Information Systems II

Proposed topics

Advertised Topic Suggestions

Please use for any request only the contact adress you can find below.

The following list contains topic suggestions for Diploma-, Bachelor-, Master theses and project studies. The specific configuration (can and) should be discussed and matched with your supervisor. Of course, the topics can be adjusted with specfic need of the supervisor in a limited context. 

Hints

Topics supervised by WIUS can serve not only the content in teaching, but also research interests and current dissertation projects. Students who want to write their thesis with companies (e.g. in the context of a working student or internship) are strongly advised to contact WIUS BEFORE signing a contract for topic coordination. In any case, own topic suggestions are to be proposed in the same form as the topics defined by the chair.

If you are interested in a thesis at the chair, please use the contact address below and refrain from direct inquiries to the staff.

A mini exposé of max. attach 1 page based on the sheet for scientific work, which particularly emphasizes the relationship to research carried out by the chair.

Agile process models in distributed software development projects

Problem

Product development in the context of software-intensive business is faced with the challenge of meeting two developments that have been going on for years. On the one hand, agile process models (e.g. Scrum) gain importance above all in software development, due to a highly dynamic environment and short product life cycles. On the other hand, increasing globalization and multidisciplinary, cooperative product development are leading to geographically distributed project environments.

These two developments present science and practice with previously insufficiently researched problems. In many aspects, agile process models are not compatible with the framework conditions of distributed project environments. For example, agile procedural models generally provide for personal, synchronous communication, while extensive documentation is dispensed with. In contrast, distributed project environments imply impersonal, asynchronous communication (e.g. via email, telephone / video conferences, ...) and a high level of documentation.

This apparent contradiction and the associated incompatibility of the two developments often leads to tailoring measures in practice, so that agile process models with plan-driven process models are integrated into so-called "hybrid process models". However, the role of the RE, especially with regard to the task area of requirement prioritization, in agile software development projects, in particular in the context of distributed project changes, has so far been insufficiently researched. Additional research work at the conceptual level (e.g. by developing a methodological approach) is sometimes necessary.

Goals of the work

Against the background of the problem presented, the following (relatively general) topics for theses arise, which can make an essential contribution to research. These can be adjusted or refined in consultation with the supervisor.

  • Analysis of existing methodological approaches for prioritizing requirements in distributed, agile software development projects in science and / or practice.
  • Analysis of existing organizational approaches (roles, ...) in the context of agile software development projects in distributed project environments.
  • Analysis of existing software support (tools, ...) for prioritizing requirements in agile, distributed software development projects.
  • Collection of requirements for software-based requirement prioritization in agile, distributed software development projects.
  • <Own topic proposal for working on the indicated problem>
Advisor

Kleophas Model

Problem

Cloud Manufacturing, created in 2009, is a new paradigm in production. The spread of the Internet, the variety of information technology (IT), an increasingly complex production and the trend change from mass production to mass individualization are examples of current challenges. The challenges are summed up by the so-called goal of “TQCSEFK” (i.e., fastest time-to-market, highest quality, lowest cost, best service, cleanest environment, greatest flexibility, and high knowledge). Cloud Manufacturing tries to promote and implement this. The high-level idea of ​​this paradigm is the interaction of the integration of distributed resources (i.e. production resources or production skills) and the distribution of integrated resources. In other words, within cloud manufacturing, these distributed resources are encapsulated in cloud services and managed in a centralized manner.

The aim of the work

The aim of this thesis is to carry out an empirical study on the topic of cloud manufacturing. This is supposed to relate the current state of perception. of this production paradigm are evaluated. Is Cloud Manufacturing Known? What challenges are currently associated with this topic? What potential is seen? This study is intended to be carried out within a specific industry.

Advisor

Robert Henzel

Problem

To develop complex end-to-end solutions such as B. Predictive maintenance has to be brought together by different actors, because no single company brings the variety of resources and skills to provide such a complex digital service in the context of the industrial Internet of Things (iIoT). That is why companies like to network with partners and traditional rigid supplier relationships are increasingly being replaced by flexible partnership relationships. The partners even become an important "resource" because the internal skills of a company can be expanded through the external skills of the partners. The networking of partners can be achieved through the use of a common technology such as B. an iIoT platform (such as MindSphere, Adamos, Forcam Force, MS Azure etc). A platform provider can achieve competitive advantages by successfully managing its partners (or even by managing the partners among themselves). Nevertheless, partner management as a research area can speak of a research gap because little research has so far existed on this concept (especially in the context of iIoT).

The aim of the work

The work comprises two goals. The first goal of the work includes conceptual work around partner management in the iIoT context. With the help of such scientific methods as. For example, systematic literature analysis or bibliometry (the methodology is taught by the supervisor), all existing approaches to partner management should be brought together. As a result, a taxonomy or another overview is conceivable, which represents all important approaches and features of partner management, clustered and arranged in a meaningful way.The second goal is to process the current actual state of the practice. In this part of the work, e.g. B. With the help of the analysis (keyword mining) of job advertisements for partner management or ecosystem management, a requirement profile can be created in order to derive relevant characteristics of partner management.The end result of the work would be a comparison of the two partial results from science and practice in order to obtain a holistic and up-to-date understanding of the concept of partner management in platform-based iIoT ecosystems.

Supervisor
Dimitri Petrik

Problem

Cloud Manufacturing, created in 2009, is a new paradigm in production. The spread of the Internet, the variety of information technology (IT), an increasingly complex production and the trend change from mass production to mass individualization are examples of current challenges. The challenges are summed up by the so-called goal of “TQCSEFK” (i.e., fastest time-to-market, highest quality, lowest cost, best service, cleanest environment, greatest flexibility, and high knowledge). Cloud Manufacturing tries to promote and implement this. The high-level idea of ​​this paradigm is the interaction of the integration of distributed resources (i.e. production resources or production skills) and the distribution of integrated resources. In other words, within cloud manufacturing, these distributed resources are encapsulated in cloud services and managed in a centralized manner.

The aim of the work

Cloud manufacturing as a new production theory requires an industry-distinct analysis. Every industry is different, given by legal regulations, manufacturing processes or structural events. In line with this finding, a theoretical and empirical analysis has already been carried out in the agricultural machinery industry for cloud manufacturing.

The aim of this work is to further enrich the existing knowledge about qualitative research methods from the social sciences. These can be, for example, expert interviews or group discussions. The resulting artifact is intended to confirm the previous findings: (a), (b) refute and / or (c) enrich by supplementation. The choice of the qualitative research method is not clearly defined, but must fit into the context of the work.

Supervisor
Robert Henzel

Problem

Providers of platform-based software solutions e.g. B. in the field of operating systems such as Microsoft (desktop) or Google and Apple (mobile) have proven in recent years that open platforms ,based on external solutions from third-party providers, prevail in the competition. These examples show that these companies, as focal companies in a platform-based network, have managed to design their partner network through various control and governance measures in such a way that it is attractive for external third-party providers (complementors). These platform providers have designed their interfaces (APIs) so that their platform solutions have become attractive for external developers.

Such ecosystems consisting of interfaces are interoperable and connect many actors. On the basis of network theory, such ecosystems can be "mapped" and visualized in order to gain new insights into their dynamics, influencing factors and strategies for the ecosystem participants.

The aim of the work

The aim of the work is to familiarize yourself with the design of the interfaces (API Management, API Lifecycle, API Evolution, API Economy) and to model them using an exemplary platform-based ecosystem (e.g. MindSphere, Azure, Predix, Cumulocity, Adamos). Then, recommendations for actions for platform providers or complementary actors in the ecosystem can be derived.

The sector examined (e.g. networked production, networked logistics, networked automobiles, ...) is determined in consultation with the supervisor.

Supervisor
Dimitri Petrik

 

Problem

Platform-based ecosystems in the IoT or industrial IoT (iIoT) benefit greatly from network effects. To generate this, the platform providers must have enough partners and intermediate providers to attract their IoT platform. In practice, however, it has so far been shown that hardly any IoT platform provider has managed to gain many partners and build large ecosystems, as is the case in business-to-consumer (Apple iOS, Google Android) or corporate software (SAP, Microsoft, Oracle) is the case. The empirical analysis of the reasons for this current state, as well as the empirical mapping of the IoT platforms themselves, therefore form promising research directions.

Goals of work

The aim of this work is based on the use of empirical methods to determine new knowledge or the evaluation (confirmation) of existing knowledge (e.g. from the literature). Various "tools" and research methods can be used for this:
- Interviews (semi-structured, critical events, basic theory)
- evaluation sheets- Document analysis (internet sources, repositories, press releases, slideshare)
- Social media analytics (data from Twitter, Xing, LinkedIn, Stackoverflow)

In an evaluation, operationalization (with the support of the supervisor) is desired and forms the core of this thesis. Initial data and previous studies can also be provided by the supervisor.

Supervisor
Dimitri Petrik

 

Problem

Cloud Manufacturing, created in 2009, is a new paradigm in production. The spread of the Internet, the variety of information technology (IT), an increasingly complex production and the trend change from mass production to mass individualization are examples of current challenges. The challenges are summed up by the so-called goal of “TQCSEFK” (i.e., fastest time-to-market, highest quality, lowest cost, best service, cleanest environment, greatest flexibility, and high knowledge). Cloud Manufacturing tries to promote and implement this. The high-level idea of ​​this paradigm is the interaction of the integration of distributed resources (i.e. production resources or production skills) and the distribution of integrated resources. In other words, within cloud manufacturing, these distributed resources are encapsulated in cloud services and managed in a centralized manner.

The aim of the work

A vignette study design is to be developed based on an existing conceptual trust model. The so-called vignette study is a social science research method. A vignette in psychological and sociological experiments represents a hypothetical situation to which the research participants react in order to show their perceptions, values, social norms or impressions of events.

On the one hand, the goal is to operationalize a conceptual model for a vignette study (a) and to develop a corresponding general procedure model (b). The resulting artifact in (a) is thus the exemplary application of the developed procedure from (b). In other words, the procedure to be described from (b) is a recommended general procedure when designing the transfer of conceptual models into a specific vignette study design. This should then be processed for an existing model (a) and implemented as a prototype.

Supervisor
Robert Henzel

 

Problem

The Chair of Business Administration and Information Systems II conducts research in the area of Quality Function Deployment (QFD) for customer-oriented development of products, services and software. For the application of this method, the software easyQFD was developed at the chair, which was further developed in the further course as a browser-based application openQFD. This program requires some adjustments in the context of agile development methods for distributed software development.

The aim of the work

As part of the final thesis advertised, the existing software is to be maintained to a lesser extent and expanded to a greater extent. The desired additions include, in particular, the improved possibility of using the software asynchronously in distributed teams. The focus is on collaborative work by stakeholders and groups. In addition to creating the concept, use cases should be developed on the basis of various scenarios and implemented in the tool.

Recommended requirements
Java, basic knowledge of JSF / JSP desirable

Supervisor
Dr. Sixten Schockert

Problem

IoT platforms from providers such as Microsoft (Azure), Amazon (AWS IoT), Siemens (MindSphere), General Electric (Predix) or SAP (Leonardo) have the goal of interoperability of various systems (e.g. cyber-physical systems and Increase business software such as ERP, CRM, MES) and connect them to each other. Such platforms (if open) are modularly expandable and can be supplemented by modular code extensions (apps). Such apps can be developed by a community or third party providers. External developers usually use certain platform resources (such as interfaces, documentation, etc.) to develop apps or connect machines based on this platform. The quality of these resources plays a major role in the satisfaction of users (or developers) with the platform and can also influence the developer's decision to recommend this platform to his customers or not.

Goals of work

In this work, the developer forums and other "Repositories" are searched for certain critical events in order to make statements about the implementation quality of a platform. The corresponding repositories can be searched for positive and negative events and opinions in order to make statements about the implementation quality of a platform. Unpleasant experience when using the interfaces (not performant enough) or usability (menus that are too long) can be understood as negative critical events. These events are to be analyzed in technical repositories and exchange forums for developers for IoT with the help of a so-called "repository mining" process.

The method used (taught by the supervisor to the student) is the so-called Critical Incident Technique by Flanagan (1954). In addition to observations and interviews, this methodology can also be applied to secondary data (e.g. from the repositories or forums).

Especially in the new domains such as IoT or industrial IoT, experience and knowledge about relevant features and quality features of a platform are still lacking. This gap is to be closed with the help of the mining process of repositories based on the methodology of the critical events. The results of this work are open and can be varied. In consultation with the supervisor, relevant events can be summarized in meaningful clusters or even new theories can be generated. For programming enthusiasts, tool development for automating the mining process is also conceivable (but not absolutely necessary).

The work can also focus on data science. That means extracting the content from the repositories and evaluating them can be analyzed using data science methods and corresponding libraries in Python or R. A sentiment analysis can be mentioned as an example to interpret the critical events from repositories. It is also conceivable to develop an automated process for extracting the critical events (i.e. the development of a scrapper with the addition of sentiment analysis).

Supervisor
Dimitri Petrik

 

Problem

In recent years, successful platform-based start-ups such as Airbnb, Uber have popularized the term platform. They play an important role in the digitalization of industry (Industry 4.0, Internet of Things). Apps and services bring added value for customers and are met via networked smart products. App developers or external service providers are complementary third-party providers that add value, on which the provider of a platform relies in order to cope with a multitude of use cases.

The platforms used in the industrial context are mostly cloud-based development platforms and (similar to Android, iOS or Windows) enable the development of applications by third parties. In order to enable the third party providers, a platform provider must provide various resources. These resources are used by the third party providers. Developer portals represent an exemplary and very important resource and should be researched in the context of this work..

The aim of the work

The primary goal of the work is to examine important design criteria and quality features of developer portals for software development platforms or industrial IoT platforms (e.g. Siemens Mindsphere, GE Predix, SAP Leonardo, Cumulocity, MS Azure, AWS IoT etc.). As a scientific methodology such. B. A systematic literature analysis, as well as a document analysis to identify the potential quality features of a developer portal, as well as the common solution features from practice and to represent them systematically.

Supervisor
Dimitri Petrik

 

Problem

In the context of technological and socio-economic changes, which are often described with the terms digitalization and digital transformation, there is also a change in product development. Hybrid products, e.g. so-called embedded systems, product service systems present product development with challenges that have to be considered beyond the specifics of the respective domain. It is expected that the ever increasing convergence of these development areas will increase these challenges in the future.

The aim of the work

Various problems can be dealt with in this context. Possible key questions and motives for investigations in this research area can be, for example (not exhaustive):

  • What practical challenges do companies in software-intensive business face when developing products?*
  • How does practice assess the importance of the challenges in product development in software-intensive business that are considered in the scientific literature? *
  • Which methods do companies use for product development in software-intensive business or for the development of hybrid products?*
  • Importance of methods such as design thinking and quality function deployment for product development in software-intensive business*
  • Practical case studies in the context of product development in software-intensive business
  • Evaluation of agile development models such as Scrum and Kanban for product development in software-intensive business
  • Evaluation of approaches for the integration of customers and stakeholders in product development (co-creation approaches, crowd sourcing ..)

Topics marked with (*) are particularly suitable for processing in the context of a practical thesis. Your own topic suggestions in this context are also welcome.

Supervisor
Felix Schönhofen

 

Problem

With its certification to the Certified Professional Requirements Engineering Foundation Level (CPRE-FL), the International Requirements Engineering Board (IREB) has created a recognized standard for basic content of RE. Quality Function Deployment (QFD) as a method for customer-oriented product planning has also achieved international standardization with the ISO standard 16355-1. To study is how these two specifications fit together, and whether there are contradictions or in which areas they match or complement each other well.

The aim of the work

Identify differences in the content of ISO 16355 and CPRE and possible reasons for it

Supervisor
Dr. Sixten Schockert

 

Cooperation and practical theses

Problem

Current sensors in digitized devices and assets are able to generate very large amounts of data. When monitoring the spindle of a machine tool, so-called high-frequency data are also generated. The storage and processing of such high-frequency data pose a problem for traditional databases. Against this background, the use of so-called time series databases (TSDB) makes sense because TSDBs have the goal of storing time series data efficiently. Most sensor data or stock exchange data is usually time series data.

The aim of the work

The aim of the work includes the design and implementation of a benchmark for different time series databases available on the market. There are currently a large number of providers on the market for time series databases and the selection of a correct TSDB solution is becoming an increasing problem. In order to support the selection of a suitable TSDB, effective benchmark concepts are to be conceived with the help of the supervisor and evaluated by application to exemplary databases (such as InfluxDB, HarperDB, crateDB). The benchmarks should primarily be embedded in the IoT context, whereby the requirement is not absolutely necessary.

The support includes close coordination and help with the design of the benchmark. In addition, the supervisor helps to obtain appropriate licenses for the databases to be evaluated and provides the necessary hardware (e.g. a Raspberry Pi and sensors) for the test purposes.

Supervisor
Dimitri Petrik

 

Registration of a thesis

 

Contact adress

To the top of the page