AIO - R. van der Meij
Innovation for knowledge consolidation

Background

Early choices made while planning a geotechnical project can have very large consequences later on. This early stage of a civil project is dominated by urban and rural planners, not by geotechnical engineers. These planners make their decisions based on their collective knowledge. The geotechnical engineer usually joins the project when the plans are rather well defined and only details can be changed. Because the geotechnical engineer has so little room for the changes, they are often expensive, inconvenient and time consuming. This can be avoided by making the right geotechnical decisions in an early stage of the project.
The goal of this project is to make the knowledge of the geotechnical engineers easily accessible for planners. Systems need to be built to capture the collective knowledge of the geotechnical engineers so it can be used in a planning phase. If such a system is made, a planner will have much more knowledge of the financial consequences of his decisions. On top of that, engineers don’t need to make adjustments because of bad decisions made early in the project, but can focus solely on the design.

Project plan

In the first two years of the project (2007 to 2008), many small cases will be worked out in order to get experience with the combination of geotechnical engineering and artificial intelligence techniques. In the third year, a large, substantial and influential project will be performed as the main core of the PhD study. In the fourth and last year (2010), the results will be evaluated, deepened, and written.

Project results

In the first two years, the following pilots have been performed:
- Modeling the stress-strain relationship of peat using an artificial neural network
- Predicting horizontal deformations using an artificial neural network (or fuzzy logic)
- Quickly designing a flood protection embankment that adheres to the Dutch VTV

In 2009 and 2010, the following projects are planned:
- Finding the stability of an embankment using a Genetic Algorithm
- Quickly calculating an emergency measure for a flood protection embankment
If you want to know more about these projects, you can check the publications on the right, or contact the autor at Raymond.vanderMeij@Deltares.nl.

Chair

The promoter is prof: F. van Tol.

Comittee

•  F. van Tol (chairman) TU Delft
•  P. Vermeer University of Stuttgart
•  S. van Baars TU Delft
•  A. van der Stoel Crux engineering
•  M. Visschedijk GeoDelft
•  H. Sellmeijer GeoDelft