TANGRAM PROJECT

Model-based integration and testing of complex high-tech products

The Tangram project aims at a significant reduction of lead time and cost in the integration and test phase of complex high-tech products. At the same time the product quality should be maintained or improved.

The main reason to initiate this project is fairly simple.

The performance demands on high-tech products keep on growing; they should be faster, more accurate, their uptime should be increased, etc. The business demands on these products keep on growing as well; the time to get such products to the market is getting shorter and the same goes for the period in which the return on investment can be obtained.
So while engineers have to do their utmost to deliver technology that sometimes has not been invented yet, the market dictates them to do it faster, cheaper and better. This challenge is never more present than when system parts from different projects and from different disciplines (optics, mechanics, electronics and software) have to be integrated and tested.
The combination of this 'faster, cheaper and better' issue and integrating multi-disciplinary state of the art technology, gives ample reason to want a breakthrough. It gives ample reason to want Tangram.

Model-based integration and test to provide a breakthrough

The solutions are less simple. Nevertheless there is a firm belief in the Tangram hypothesis that model-based integration and test will result in significant improvements. Tangram foresees research and development along four Lines of Attention (LoA) that will be constantly challenged by a real life industrial case: a wafer scanner at ASML.

Lines of Attention

  • The LoA with focus on strategy will develop models of integration- and test processes featuring cost, parts to integrate, combinations to test, time to spend and product quality to achieve, as well as methods for test selection.
  • The LoA that concentrates on test generation & execution will deliver model-based methods and tools for generating, executing and evaluating testcases as well as model-based simulations for parts that are not yet available.
  • The LoA that aims at model-based diagnosis will deliver methods and tools that are capable of diagnosing the internals of a system by monitoring its inputs and outputs.
  • The infrastructure LoA will deliver an environment that is capable of interconnecting the models and tools that result from the other LoA's.

Another view on where and how we want to apply models is depicted in the picture below

Models

Partners

Tangram has teamed up the expertise and competence required to establish a breakthrough. Three universities: Delft University of Technology; Eindhoven University of Technology and Radboud University Nijmegen. The institutes ESI and TNO-TPD. The industrial partner Science & Technology and finally ASML as Carrying Industrial Partner.
The project is partly subsidized by the Dutch Ministry of Economic Affairs.

Overview of the Tangram publications