Making sense of big data to improve the nation’s defence, security and prosperity

Dstl challenges data scientists to solve real world challenges in first public competition

The Defence Science and Technology Laboratory (Dstl) has today announced the launch of the Data Science Challenge. The challenge is designed to bring the brightest minds in data science together to solve real-world problems. The first challenges – detecting and classifying vehicles from aerial imagery and the classification of documents by themes– are now open to entrants, with each challenge boasting a total prize fund of £40,000.

The Data Science Challenge is part of a wider programme set out in the Defence Innovation Initiative that aims to build an open innovation ‘ecosystem’, harnessing the talents of individuals, academia and industry to develop new approaches to complex problems. The Data Science Challenge is piloting new ways of working including the use of crowdsourcing to engage the data science community to develop cutting edge solutions to Defence and Security problems.

The Data Science Challenge includes two distinct problems that will test the participants’ ability to mine large unstructured datasets to extract useful information:

Being able to automatically detect and categorise vehicles in aerial imagery will dramatically improve how quickly we can assess and identify them. This challenge asks participants to detect and classify vehicles such as buses, cars and motorbikes, from a set of aerial images.

Analysing data in documents such as media reports can provide a better understanding of a potential crisis situation, growing instability in a particular region or specific theme such as terrorism. Using news material, this challenge asks participants to predict topic tags for classifying unseen reports so that they can be used to improve awareness and understanding.

Minister for Defence Procurement Harriett Baldwin MP said: