Afstuderen of stage op het snijvlak van universiteit en bedrijfsleven

Axini is een spin-off van de Universiteit Twente. We passen academische technologie toe in de praktijk, in samenwerking met universiteiten. Dit maakt dat we spectaculair betere resultaten boeken dan onze concurrenten en het biedt interessante mogelijkheden voor afstudeer- en stageopdrachten.

We verwelkomen universitaire studenten informatica met een kritische en pragmatische blik. Bij ons vind je een prikkelende omgeving, goede begeleiding in combinatie met veel vrijheid en toepassing van je werk op cases van klanten.

Heb je al een specifieke opdracht voor ogen, laat het ons horen. Ben je nog zoekende, bekijk dan een van de bestaande opdrachten onderaan of doe inspiratie op uit de volgende onderwerpen die bij ons spelen:

Ontwikkeling van domeinspecifieke talen in Ruby
Praktische toepasbaarheid van modelleertalen als Promela, Alloy, executable UML en SBVR
Uitbreiding van ons test-framework, gebaseerd op Ruby on Rails
Gebruik van constraint solvers, theorem provers en model-checkers
Toepasbaarheid van academische test tooling als TorX, TGV en Alloy
Meten en analyseren van testcoverage op basis van modellen en implementatie
Visualisatie van modellen en testresultaten
Gebruik van AI-technieken, zoals evolutionaire algoritmen, voor het genereren en optimaliseren van testverzamelingen
Geautomatiseerd testen in het algemeen en model-gebaseerd testen in het bijzonder

Bespreek de mogelijkheden met Machiel van der Bijl: 06-16426332 of vdbijl [at] axini.com.

Openstaande opdrachten

Project: Alloy for test case generation in the real world

In this project you research the applicability of Alloy (alloy.mit.edu) for test case generation with TorX. Alloy is an interesting modeling language because data-intensive applications can be modeled easily. This is for example very important for the financial market, because many of the systems in this area are data-intensive (mortgage systems, investment systems, etc.).

You will do this project with Axini, a spin-off of the University of Twente specialized in automated testing of complex systems. The use of Alloy and the Alloy Analyzer looks like a promising extension of the current tool chain (TorX with Promela, dedicated tooling in Ruby and Rails).

We identify the following milestones in the project:

Literature study on test case generation with Alloy. Part of this will be the TestEra tool set, a tool set for model-based testing of Java applications
Modeling of a case study of one of our clients. An important aspect will be the readability of the model
Development of a test case generator that analyzes Alloy models and generates test cases
Evaluation of the use of Alloy for test case generation. Aspects of the evaluation are the achieved coverage and the readability of the model

It is possible, and we highly recommend, to do part of the project on site at one of our clients, for example to execute the test cases on a real system under test.

We are looking for whip smart, pro-active students that have an interest in testing and tool development. We offer a mind tickling graduation environment right between science and practice. If you have ideas of your own to enhance or change the project, let us know! For more information contact Machiel van der Bijl, vdbijl [at] axini.com (06-16426332)

Project: coverage-driven test case generation in Ruby

Model-driven testing allows one to derive tests automatically from a specification. Since testing is inherently incomplete (“testing can only show the presence of bugs, not their absence”), test selection is of vital importance: one aims at executing a test suite with maximum impact at minimum cost.

The idea in this assignment is to implement a notion of coverage of functionality in terms of the specification. Based on this we can define a notion of distance on test cases. Once we have these two concepts in operation, we can compare test cases with one another and can define criteria why we think that one is better than the other based on coverage information.

You will do this project with Axini, a spin-off of the University of Twente specialized in automated testing of complex systems. The application of coverage and coverage driven test generation looks like a promising extension of the current tool chain (TorX with Promela, dedicated tooling in Ruby and Rails).

We identify the following milestones in the project:

Literature study on model-based testing and coverage
Definition of a notion of coverage and a notion of distance between tests
Design and development of an algorithm to compute coverage of a specification and to drive test case generation based on this information
Case study of the algorithm with one of Axini’s customers