Incident modelling and information fusion solutions expert, Riskaware, has just delivered a new prototype counter-drone capability, EDITTS, to the Defence Science Technology Laboratory (Dstl). This latest solution was developed under an innovation award from the Defence and Security Accelerator (DASA). 

The Enhanced Drone Identification and Target Tracking System, (EDITTS) was conceived in response to the DASA Countering Small Drones phase 1 directed call in 2019. This competition aimed to find technology advances that could improve the ability of the UK defence and security community to counter an adversary’s use of UAS while minimising collateral damage. 

By exploiting advanced information fusion and machine learning methods, including neural networks and particle filters, our solution effectively addresses several key challenges within the overall programme goals. Not only can the EDITTS provide improved tracking of UAS (Unmanned Aerial System) targets, but it also demonstrates the capability to distinguish between different types of drone target – such as fixed wing or rotary wing drones – and between drones and “non-drone” targets, such as birds. 

The EDITTS system is designed to minimise distractions and false alarms, to allow human operators and counter-UAS systems to focus on real threats. This will enable faster identification and more effective countering of UAS threats within civilian settings, such as airports and future combat environments. 

Martyn Bull, Technical Director and Coordinator of Riskaware’s Innovation Hub stated: “We are very grateful to DASA for funding this work, which has allowed Riskaware to participate in this important new marketplace. We are always excited to find new ways of using our modelling and algorithmic expertise to add value within different sectors and to address important challenges within defence and security.” 

During the project, the Riskaware team collaborated with Prof. Simon Maskell of the University of Liverpool, to develop and test groundbreaking features of the solution. This included  implementing established machine learning methods for processing target imagery, and devising new approaches to identify the type of target using the target dynamics. 

The Riskaware team also utilised Dstl’s open-source Stone Soup – a software project designed to provide the target tracking community with a framework for development and testing of tracking algorithms – throughout the programme. This helped enable the development and evaluation of key algorithms, and provided an excellent platform for developing the concepts that underpin EDITTS.

This work builds upon Riskaware’s existing experience in developing neural networks, particle filters and similar algorithms to provide actionable intelligence to address a range of other defence and non-defence challenges, including disease outbreak detection, cyber-security and CBRN source estimation.

Martyn Bull stated “We are proud to collaborate with such great partners, and to apply our collective expertise to find new solutions and insights that can help people tackle these kinds of real-world challenges.”