Faculty of Science and Technology

Dr Plamen Angelov on a visit to the Henry Ford Museum in the US

Patented software developed by Lancaster University for use in industry has been licensed by the US multinational Ford.

The software - called EST or Evolving Systems Toolbox - is capable of self-learning and so requires minimal human intervention. It is being used to monitor the "health" of vehicles and machines that produce vehicles and to monitor reactions such as how drivers press pedals.

The technique is based on the latest academic research and has applications in everything from transport and logistics to defence and the internet. It was developed by Dr Plamen Angelov, Reader in Computational Intelligence at the School of Computing and Communications at InfoLab21 - who is an expert on evolving intelligence systems - and his team of students and researchers.

Other applications based on the research include self-learning virtual or "smart" sensors in industry which are able to self-calibrate and so reduce maintenance costs. An oil refinery in Tenerife owned by the Spanish industrial group CEPSA, is using these self-learning "smart" sensors to estimate the yield and quality of different oil products in real-time.

Dr Jose Juan Macias Hernandez, the Process Engineering Department Manager of CEPSA, said the application of this technology in just one area - such as atmospheric distillation - could mean savings of tens of thousands of Euros a year.

"It is of vital importance for the CEPSA refinery to know in advance the yield and quality of products that will be obtained from certain input streams. The application and development of this technology opens a whole new exciting horizon in the petrochemical industry."

Dr Angelov has also worked with one of the world's largest chemical companies, the Dow Chemical Company in Texas, to develop self-learning "smart" sensors which are used to reduce shutdown and improve product quality.

Dr Arthur Kordon, Data Mining and Modelling leader at the Advanced Analytics Group of the Dow Chemical Company said that the total impact of "smart" sensors had saved his company an estimated $30m a year.

Further applications which are at the testing stage include: helping medical doctors in dosing Warfarin drugs to patients with DVT (deep vein thrombosis) and classification of spectroscopy data for biomarker identification (jointly with Dr Frank Martin of the Lancaster Environment Centre and funded by Unilever). The Lancaster University spinout company Entelsensys Ltd is developing sensor systems based on applications of this technology in other industrial sectors.

In recognition of the growing importance of this research area, the IEEE - the world leading organisation in this area - has established a Technical Committee on Evolving Intelligent Systems and organises annual conferences chaired by Dr Angelov, who is also editor-in-chief of the journal "Evolving Systems" published by Springer.

Thu 05 May 2011