Your browser is no longer supported

For the best possible experience using our website we recommend you upgrade to a newer version or another browser.

Your browser appears to have cookies disabled. For the best experience of this website, please enable cookies in your browser

We'll assume we have your consent to use cookies, for example so you won't need to log in each time you visit our site.
Learn more

'Brain damaged worm' predicts floods

ACADEMICS AT Leeds University have developed an artificially intelligent computer system which they claim will revolutionise flood forecasting - despite having the thinking capacity of a 'brain damaged worm'.

The 'intelligent' system - made up of a network of 24 artificial neurons - runs on conventional hardware and software but mimics the working of a biological brain by learning through experience to predict flood levels.

Research to produce and test the £90,000 neural net has been funded by the Ministry of Agriculture, Fisheries & Food - the ministry responsible for flood and coastal defences. Project leader and professor of human geography at Leeds, Stan Openshaw, claimed that early comparisons showed the net to be consistently better at forecasting than either conventional hydrodynamic models or time series forecasting.

'Neural nets have the advantage of flexibility, ease of construction and maintenance, and speed of application. At a time when the weather patterns appear to be changing and becoming more extreme, this flexibility is going to become increasingly important,' he said.

In a conventional mathematical or statistical model an equation has to be invented that maps the input onto the output forecast - a difficult process which relies on expert knowledge of the kinetic processes at work. But the artificial neural network - a type of regression model - is able to derive the relationship between the input and output by trial and error.

In comparison with the 24 neurons in the artificial net, the human brain is thought to contain 1011 neurons, while a worm's brain contains around 1,000. However, Openshaw claims the system is still intelligent enough to adapt to new data and catchment area behaviour over time, and could even recognise if it makes a mistake. Researchers at Leeds hope to develop a PC portable software package this year that could be used by any engineer.

But news of the development was met with scepticism by the Environment Agency - the body charged with forecasting floods. MAFF/EA research advisory committee chairman Edmund Penning-Rowsell commented: 'I believe more money should be spent on flood warning research but it's not just the technical aspects which we should look at. During the Easter flooding human behaviour was particularly important as well.'

Matthew Jones

Have your say

You must sign in to make a comment

Please remember that the submission of any material is governed by our Terms and Conditions and by submitting material you confirm your agreement to these Terms and Conditions. Please note comments made online may also be published in the print edition of New Civil Engineer. Links may be included in your comments but HTML is not permitted.