Methods

Modeling is the key for analysis and design; and purposeful abstraction by means of an appropriate model is the tool. I rely both on generic approaches like domain driven design and business process modeling, as well as on more specific methods like entity-relationship data modeling, object-oriented modeling, and domain-specific mathematical models; e.g., for dynamic systems and automata. My modeling experience ranges from integrated circuits and communication systems, multi-domain simulation models of electric vehicles and hybrids to business processes in energy and healthcare; it also covers the corresponding implementation models in languages like ER, UML and BPMN.

Data analysis is all the hype currently. First, because many relations in complex systems are no longer in plain sight but must be derived from thousands or millions of observations; second, because data is abundantly available and processing is cheap. However, machine learning and AI are not a panacea; in particular, they cannot substitute a sound understanding of the domain. To derive insights from data, I start with proven statistical methods like correlation- and regression-analysis and, if necessary, employ advanced techniques like model order selection.

The knowledge available in the scholarly literature doubles every ten to fifteen years; the amount of general written information doubles almost every year. However, the growth rate of relevant knowledge is far lower. Thus, it is essential to quickly filter out the noise and chase the pieces of relevant information given very few leads. The open internet is my primary resource, especially for all things related to IT. Yet, I also have ample experience researching in specialized databases, tracing chains of references, and spotting the essential arguments. In addition, I value a well-sorted collection of standard textbooks, because core ideas do not get out of fashion.

I prefer to compile the results of my work as coherent written documentation – in German or English. To this end, I spent lots of time practicing technical writing, Slides are an effective means to support a presentation. Yet, I do not consider them as sufficient technical documentation on their own, because they often do not allow for reconstruction of design decisions later on. A consistent, logically sound, and readable exposition also significantly raises the probability of success in both public and private tenders, and it increases customer confidence.

Agile methods help customers and teams to jointly discover and learn about the actually needed system in a risk-optimized way. Especially for tackling new requirements, new technologies and new markets, I do recommend an agile approach. Scrum and the Scaled Agile Framework have been mainstay for my teams during the last years. However, applying an agile method by the book is not sufficient; most important is trust: the trust management places in their teams, and the trust amongst team members.

Workshops are the method of choice for customers and teams to jointly discover domain models and comprehend requirements; and meetings are everywhere also at smaller scale. However, meeting efficiency is always of concern. For a workshop to be efficient, with results that justify the large expense in personell, professional preparation is often warranted. I underwent training as workshop facilitator at my time at Robert Bosch GmbH, and have conducted many internal and external workshops ever since, with focus on process optimization, requirements development, project setup and review.

Developing a complex software-intensive system is most often not a textbook project with a well-defined endpoint; rather, an initial system nucleus must be adapted to changing circumstances and the growing understanding within the organization. Project management methods are important to keep development on track, even though the textbook systems project does not exist. For me, project management is risk management: a framework to quickly and adequately react upon new insights from agile development, and a method to take decisions that require trading off resources, quality, and scope. I underwent trainings both in conventional and agile project management at my former employers, and I collected ample practical experience in every development project since.

People management was one of my core duties during the past ten years. It is people who make or break a system, both during development and operation. The one most important asset of a world-class team is a shared mental model of the system the team members are building or operating: a shared understanding of the strategic goals, core processes and fundamental structure of the system, which outlasts documentation, management and individual system components. Thus, people management for me is all about building the trust and providing the qualifications that allow a team to develop an appropriate shared mental model and to own it.