IBM Faculty Award for two Groningen scientists

30 November 2011

Prof. Lambert Schomaker and Prof. Rien Herber each received an IBM Faculty Award on 29 November. These awards are presented annually by IBM for extraordinary activities in the field of IT research and innovation. Herber and Schomaker are the only Dutch researchers to receive a Faculty Award this year.

Schomaker

Professor Schomaker, professor of Artificial Intelligence, has won the award for developing the ‘Monk’ system, which enables searches in handwritten historical documents. The system could be developed because word images of these documents are currently available via internet, accompanied by the correct related text. With their help, a computer gradually learns to read the illustrations.

Schomaker: ‘What is special about this approach is that there is no question of a laboratory phase and a separate operational stage in the real world. What is happening is continuous learning of material supplied by a number of Dutch archives. This approach makes special, new demands of computer systems.’

New historical documents that can be processed come from the Groningen Archives and the Louvain City Archive.

Herber

Professor Herber, professor of Geo-energy, received the award for research on the prediction of tremors related to gas extraction. Because the pressure in gas fields declines the more gas is extracted, the reservoir ‘readjusts’ from time to time. That is experienced on the earth’s surface as a tremor that can be clearly felt and registered. The question is, is it possible to predict when underground tremors that can be felt on the surface are going to happen?

Herber: ‘In order to answer that question, we have to have an accurate image of the dynamics underground at about 3-4 km deep. To that end we not only need to measure the tremors that are already felt anyway, but also the very weak ones, which are usually lost in the background noise and which pass unnoticed by people. The aim of my project is to measure these very minor tremors with special sensors that are buried at a depth of about 200 metres. You could regard them as a kind of warning signal. We can then try to predict the behaviour of a virtually exhausted gas field. What is new is that this does not depend on statistics but on actual measurements.’