Freitag, 27. Januar 2017

Design Science

Design Science asks the question "How to design for a certain context?".

At the moment, there seem to be several papers who define the concept. One early source is the book of H. Simon "Design of the Artificial" from 1968. Another interesting source is Prof. Roel Wieringa's work at the University of Twente who defines Design Science as a primarily constructive (technical) science with additional "curiosity knowledge questions":
Have a look at Fig. 2 on p. 4, in which Design Science is declared to be a 4-step research process with two constructive (technical) and two "knowledge" question steps:
  1. Problem identification (knowledge question)
  2. Solution design (constructive question) 
  3. Test the design on validity to the context (knowledge question, often empirical or experimental)
  4. Implementation of the design (constructive step)
Thus, Wieringa defines Design Science as a mixture of constructive (technical) science and curiosity (descriptive) science. 
 A typical crossover disciplines of Design Science are "Wirtschaftsinformatik" (business informatics) and "Wirtschaftsingenieurswesen" (technical economics). In the ACATECH paper of Prof. Schuh et. al., this is discussed. For instance, have a look at p. 52, Fig. 18, which depicts collaboration of engineering sciences and economic science. All mixed-colored phases are talking about Design Science.

The process of "digitization" (Digitalisierung) consists mainly of the trend that software business models replace classical product- and service-models. An important task will be in the next years to define strategies for "digital transformation", i.e., for the transformation of classic companies to software companies.

To this end, the University should develop courses about software business models, e.g.,  digital HW-SW-ecosystems with platforms and complements. Design science will play an important role for this task, because it is a crossover between classic technical science and descriptive research methodology. And as Wieringa has told us, we have to blend engineering with descriptive science.

Montag, 16. Januar 2017

Smart Systems Hub Dresden

Last week, the government of Saxony has announced in a press conference the Smart Systems Hub enabling IoT, the Dresden initiative to take part in the initiative de:hubs of BMWI.

This includes two new buildings, the Lehmannzentrum part II, collecting all IT institutes of TU Dresden, as well as a Co-Innovation Center (CIC) for the Smart Systems Hub, a new form of technology center.


The taskforce of the Smart Systems Hub is headed by Prof. Frank Fitzek (ETIT, TU Dresden). Prof. Aßmann is a member. The taskforce prepares a proposal for the BMWI. In June 2017, at the IT-Summit in Ludwigshafen, Dresden wants to join the de:hubs. 

The purpose of the Smart Systems Hub is to provide a platform to transform classic industries with IoT technology ("digitization"). Therefore, a Hub Agency will mediate external, national, and international parties with Dresden and Saxon players for IoT technology. 
  • Jungle trails through the Smart Systems Hub will be defined, on which parties can meet 
  • Common research and transfer projects will be defined.
  • Prototypes of wild ideas will be tested in the IoT makerspaces and living labs of the Smart Systems Hub. 
  • Students will be involved in the Co-Working Space.
  • Corporate partners can host a jungle trail. 

Mittwoch, 4. Januar 2017

Robots Will Change the World

Robots will change the world, that's for sure. More and more interesting applications appear. Here is a list of postings:

Watch out for more!

Donnerstag, 1. Dezember 2016

Staying Power - the Power of Software Ecosystems

The business model of "platform + complements = ecosystem" is one of the best-scaling business models of today. It has been described by M. Cusumano in 2011 in his book "Staying Power". This book clearly shows why large US companies are successful today, and why many others, like German companies, loose markets. The book has taught me why Germany has not been able to win important IT platforms over the years: compilers, databases, web browsers, operating systems, everything went to the US. There is one single exception: the SAP products were, right from the beginning, planned as a platform with a software ecosystem of consultants, solution providers and plugin complementors.

Platforms create vendor-lock-ins, i.e., the ecosystem is dependent on the platform owner. Wars ("platform wars") are fought about platforms. Platforms can create enourmous large ecosystems with millions of "complementors". 

If Germany wants to have success in IT, it should look at this business model. 

Examples for ecosystems on monolithical platforms owned by a single company

  • Apple and Google appstores - well, everybody knows
  • Atlassian ecosystem (Jira, Confluence etc.)

Examples for ecosystems on consortial platforms owned by several players

These ecosystems are more stable because if a platform owner dies, the platform still has a chance to survive:
  • Eclipse, a layered ecosystem
  • AutoSAR, an ecosystem for control software for cars
  • Genivi, an ecosystem for in-vehicle infotainment (IVI)

Additional Literature

  • Annabelle Gawer, Michael A. Cusumano. Platform leadership. How Intel, Microsoft and Cisco drive industry innovation. Harvard Business School Press
  • Michael A. Cusumano. Staying power. Six enduring principles for managing strategy and innovation in an uncertain world. Oxford University Press. 
  • Slinger Jansen, Anthony Finkelstein, Sjaak Brinkkemper. A Sense of Community: A Research Agenda for Software Ecosystems. IEEE ICSE Companion, 2009.
  • Software Ecosystems. David G. Messerschmitt, Clemens Szyperski. MIT Press

Dienstag, 22. November 2016

Immersive Robotics

Immersive Robotics makes you dive into the robot's view of the world:

Software Institute DLR

Dresden gets a Software Institute, hosted by DLR. Some links below

Smart Clothing

Handy clothes are clothes equipped with sensors and mobile phone chips, such that the mobile phone vanishes into the cloth. In about 10 years, they will be cheap to buy because the prices for sensors and highly integrated chips shrink year by year.

Here an overview of current smart clothes: