ICT with Industry

The ICT with Industry workshop is a yearly, one-week event where a multidisciplinary group of ICT researchers, mostly PhD students at one of the Dutch universities, study real-world research problems brought in by a number of high-tech, innovative ICT companies and institutions in the Netherlands. In this session we present the results of ICT with Industry 2016, consisting of four brief reports by the companies who submitted a case, in which they will explain the content and especially the outcome and results achieved during and after the workshop.



Joost Bosman

Joost Bosman (Applied Mathematics & Informatics) is a senior IT manager and expert in software/knowledge engineering at ING. He is currently involved in the transformation of the bank towards a predictive enterprise examining the possibilities of applying Knowledge Engineering, Software Engineering, formal methods for financial industry (DSL, SMT, CodeGen, Test), Availability & Performance Engineering (Queuing theory Probability & Stochastical Networks) and applied gaming / VR  improving knowledge transfer & capturing(semantic WEB, software production, maintenance and exploitation.


Title: Sustainable agility by managing IT systems entropy

The IT world is in great flux, especially in the financial industry. Various trends, driven by customer expectations on the one hand and by new technologies on the other, are coalescing in a change that is no less than revolutionary. Looking at the market expectations concerning continuous improvement and innovation, but keeping in mind that the financial industry is a multi-billion business with a huge legacy in the form of investment in its current IT infrastructure, it is necessary to rethink agility and extend it with the concept of sustainability. Sustainable Agility should help the industry to keep up to speed against acceptable costs.

A key lesson is to master complexity. The prime method for doing so is to use abstraction to break the system down into substructures, and to very carefully define and maintain their interfaces. Failure to do so will inevitably result in an increasing resistance to change in your IT system, ending up in the chaos we are currently facing.



The core challenge in this ICT with Industry project is to objectively quantify system complexity , making it possible to follow trends on how chaos develops in your system and to react timely and appropriately. Next to that, the ability to quantify complexity opens up the possibility to also express agility in a quantitative way.

In addressing these issues, we are especially interested in n-tier, multi-platform systems, since that is the basic architecture of the existing infrastructure.

In the context of the ICT with Industry workshop, the expected contribution from the ICT research community is to answer one or more of the following questions:

  •         What existing research results and insights can be brought to bear? In particular, what can we learn from previous work in the domain of IT systems complexity? Not just in the world of finance, but also in the field of telecommunication that basically faces the same phenomena. In particular, which of the current approaches address n-tier, multi-platform systems? To what degree do current approaches account for Service Oriented Architectures and/or modern style (restful) APIs?
  •         How can complexity in IT systems be made explicit and quantifiable? For instance, how can we assign complexity to computing networks (more specific ally, instances of n-tier multiplatform)? How can we quantify this complexity? What are the tresholds that once exceeded alerts that intervention is needed to decrease complexity
  •         How can agility in changing IT systems be made explicit and quantifiable? For instance, how can we define (a form of-) modular continuity across these n-tier multiplatform application, that can be used to predict (i) the effort of applying a change in specification to the mentioned system, in terms of man-hours programming or (ii) the effect on complexity as a result of such a change? How can we express the outcome in element(s) of the total cost of ownership of the mentioned system?  
  • Can we create a (numerical-) model that expresses the (dis-) order in our IT system?
  •    Can we define trends in that model?
  •    Can we define tresholds in that model that alerts/triggers intervention in mentioned system (structure)?

Wouter Poncin

Wouter has studied Computer Science & Engineering at the TU/e on the topics Software Engineering and Process Mining. After completing this study Wouter started working at SNS REAAL (currently ‘de Volksbank’) in Den Bosch. A few years Wouter worked as a programmer on the backend systems of the mobile banking apps. Currently Wouter is working in the Technology Centre of de Volksbank where new and strategic technologies are put to the test for applicability in products for our customers.


Title: Code coverage improvement of database-centric applications


Jan de Sonneville

Jan de Sonneville is CEO of Life Science Methods BV, a startup that sells microinjection devices to biological laboratories. In addition, he is a technical coordinator (consultant) for VLPB. VLPB is an association of 10 plant-breeding companies that initiates projects in the domains of e-bioscience, bioinformatics and IT-infrastructure in which the plant-breeding industry and associated knowledge institutes collaborate. He obtained his PhD at Leiden university, on "Reinventing microinjection, new microfluidic methods for cell biology", and has a background in NanoScience, Physics (master, Leiden University & TUDelft) and Electrical Engineering (bachelor, TUDelft).


Title: Fast system to query large datasets of genomic variants

Currently, data is generated for thousands of accessions with each containing tens of millions of markers. Having many (e.g. over 1000) high-density genotyped individuals available for per species will become reality for breeding companies within the coming years. The de facto standard for storing variants is in specialized compressed binary files (BCF/VCF), which are indexed to allow for random access to specific genomic positions. Although these positional queries are very fast, they do not allow for flexible and fast interrogation of the data on other features than position.

During the workshop we identified four possible solutions, of which one was chosen for further prototyping and R&D.