The Institute for Security Science and Technology (ISST) is excited to announce the recipients of the ISST Champions Fund.
The Fund provides seed funding for new activities related to security science and technology. In total three applications were funded.
Crowd dynamics
The first project, led by Dr Bill Proud based in the Institute of Shock Physics, will investigate new methods of modelling in crowd dynamics.
Crowd dynamics is an area of study that aims to model and predict how a crowd reacts in different situations. It is particularly useful in crisis training, urban planning and evacuation simulation.
In most current crowd dynamic models, the individuals in the crowd are set to react in an unthinking way – they don’t put thought into their actions. In order to obtain more realistic results, this project aims to include an individual’s experience of the environment into the model.
For example, it is known that if a fire breaks out in a hotel, many guests will seek to evacuate via the reception, as this is a part of the hotel they’ve experienced before. The project will look at how you can define this, and study it experimentally. The results will be valuable for many crowd dynamic studies in the future.
Data privacy
The second project is led by Dr Yves-Alexandre de Montjoye, who heads up the Computational Privacy Group. This timely project will help governments, businesses and individuals understand the risk of re-identification from anonymised data.
Yves-Alexandre’s group previously showed that using just four spatio-temporal points was enough to uniquely identify 95% of the individuals in a large-scale mobile phone database. Other research in this field has shown the same can be accomplished from other data sources, such as credit cards and smartphone applications.
This project will develop statistical and machine learning methods to robustly identify individuals in large-scale behavioral datasets. It will help inform people, organisations and governments of data security and privacy risks, leading to the development of stronger data protection techniques, legislation and regulation.
Hybrid computers
The final project, led by Dr John Wickerson from EEE, aims to explore further how a new type of computer chip works, and improve their potential for use in safety-critical situations.
The new chips are called hybrid, as they combine a conventional computer processor with something called reconfigurable hardware.
The conventional processor is what carries out instructions from software in a standard computer. The reconfigurable hardware allows the programmer to execute computer instructions directly in the hardware, rather than the software. This can be up to 10x faster and more energy efficient, as the hardware is reconfigured to match the task at hand. However, the conventional computer processor, executing multiple software instructions, is more flexible. Hybrid chips provide the best of both worlds.
These new chips show great potential for applications such as image processing for pedestrian recognition in a self-driving car. However, before they are used in these safety-critical environments, we need to have a deeper understanding of how they function, as the stakes are so high.
Article text (excluding photos or graphics) © Imperial College London.
Photos and graphics subject to third party copyright used with permission or © Imperial College London.
Reporter
Max Swinscow-Hall
Institute for Security Science & Technology
