Register Now - UCL Mathematical Sciences Industrial Sandpit - 26th to 28th June 2017

Registration is now open for the first UCL Mathematical Sciences Industrial Sandpit which will be held over three days from Monday 26th to Wednesday 28th June. The sandpit will take place in the South Quad Pop Up Learning Hub (near the Print Cafe) and three external partners (Department for Transport, Motorola Solutions and DSTL) seeking engagement with the maths/stats community will be presenting challenges requiring novel mathematical/statistical ideas and techniques.

Please apply to register for a place by filling out the following short form

Registration is free and includes refreshments, lunch and an invitation to a drinks reception on Monday 26th June from 5pm-7pm.

What is the sandpit about?

The purpose of the sandpit is to get groups of academics together to explore certain themes or research pathways, with the aim of formulating possible research projects to tackle each challenge. Participants will be expected to apply their knowledge and experience across discipline areas to build a research team possessing the required expertise. Researchers across all areas of mathematical and statistical research are encourage to apply and participate. Postgraduate students will also be encouraged to take part in the process and there will be some relevant researchers from other disciplines (such as physics, engineering, computer science etc.) attending by invitation. Teams will give short presentations at the end of the sandpit showcasing their conclusions and proposal ideas. After the workshop, these short research proposals are written up and submitted to each external partner for consideration. Collaborative funding opportunities can then be explored, including the possibility of matched funding from UCL for PhD studentships arising from this event.

What are the challenges?

Here is a brief summary of the challenges:

Dept for Transport: Improved and efficient detection of weapons/complex items in vehicles

DfT are looking for novel overt or covert methods to screen vehicles (possibly with occupants inside) to detect weapons/weapon parts. Proposals need to consider number of cars screened per hour, high detection rate with a low false alarm rate where false positives are easily dealt with, and limiting ionising radiation exposure. Different imaging modalities, wave scattering inside metal cavities, inverse problems, pattern recognition, machine learning, and modelling human factors are all of interest.

Motorola Solutions: Using data analytics for predictive policing

Motorola Solutions provide important communications products and services to the emergency services. They are currently interested in developing new predictive models to tackle serious violent crime (e.g. knife crime), as well as crowd analytics to identify and predict the onset of events to improve first response logistics. We expect data sets to be made available from various sources including our contacts in UCL Dept of Security and Crime Science.

DSTL: modelling trace contamination of particles

Sensitive chemical detectors are deployed at airports to detect trace amounts of explosives contamination that adheres to surfaces when bulk explosives are used to manufacture a terrorist bomb. Explosives may be commercial or home-made, and may be powdered, crystalline or plastic (bulk crystals stabilised in a malleable plastic matrix). DSTL has conducted empirical studies of the mass and particle sizes of different forms of explosives transferred to different surfaces (eg glass, metal, plastic, cardboard, textiles) from a contaminated finger in a depletion series. During such a depletion series, individual explosives crystals will fracture under the contact pressure, resulting in changes to the particle size distribution as the print number increases. DSTL want to find a mathematical model of trace contamination that accounts for the observed mass deposited and changes to the particle size distribution on different surfaces, with the longer term goal of predicting the quantities of trace contamination arising in different scenarios.

I hope you will be able to participate in this event, which will be a great opportunity to bring the wider UCL mathematical sciences community together as well as generating further engagement with industry/government. For further information please contact Dr Nicholas Ovenden (UCL Mathematics)

This page last modified 6 June, 2017 by Nicholas Ovenden