Project aims for accurate ‘premonition’ of fire

Fire fighters and academics are working on a project to keep vulnerable communities safe from fire.

The so-called PREMONITION project has been launched at Sheffield University and Sheffield Hallam University and is being carried out by experts in behavioural risk analysis, intelligent simulations and in the study of social processes.

According to Sheffield University, computer modelling techniques such as large scale agent-based modelling will enable fire fighters to draw together different strands of information, including geographical, demographical and behavioural data to build up a picture of an area and predict where fires and other emergencies might occur.

The PREMONITION simulation will reportedly enable fire fighters identify where the most vulnerable areas are, considering factors such times of the year or the day when risks are thought to be greatest.

Some of this information is already available to fire services through online sources, or from local authority records, but due to the vast amount of data it is difficult for humans to make sense of this information and combine it in real-time to support decision-making.

Further layers of detail are also being added to the computational model about the routines and behaviours of people living within particular areas, taken from previous research of residents carried out by South Yorkshire Fire and Rescue Service, to produce even more accurate results.

The primary aim of the PREMONITION intelligent system is to enable fire services to make better decisions about where to allocate resources and improve planning and fire prevention initiatives.

Dr Daniela Romano, in the Department of Computer Science at Sheffield University, is leading the project, along with Dr Dermot Breslin, from Sheffield University Management School, Dr Stephen Dobson, in Sheffield Hallam University’s Business Systems Department, and experts from South Yorkshire Fire and Rescue Service (SYFRS)

‘We live in increasingly complex social networks, with our behaviours being influenced by many interrelated factors,’ Dr Romano said in a statement.

‘Although fire services already have access to much of this information, there is no tool that can help them grasp all of the different strands and utilise the information in real-time to make decisions. This predictive model will unpack this complexity, and help manage resources and services targeted at the most vulnerable groups in our community.’

The project is funded by SYFRS, through its Stronger Safer Communities Reserve, and is initially targeted at communities in Sheffield. If successful, the programme could be made available to fire services across the country.