Royal Navy Search for Novel AI Solutions

Published Mar-12-18

Breakthrough:
A novel tool that uses artificial intelligence to identify whether marine vessels are friend or foe.

Company:
Ministry of Defense, United Kingdom

The Story:

Data and artificial intelligence offer a lot of opportunities for improving defenses. While partially autonomous and intelligent systems have been a critical part of military technology since at least World War II, advances in machine learning and artificial intelligence are expected to be game changers. Governments and defense and security organizations in numerous countries are investing heavily in AI-related defense-specific R&D efforts which include engaging with innovators outside of the defense industry.

Engaging the Crowd

In November 2017 the UK's Ministry of Defense (MOD) hosted an AI hackathon, 'Enabling Defense Decisions'. Its ambition was to explore how AI processes such as neural networks, machine learning and deep learning could enable better decision-making and meet defense objectives more efficiently.

Approximately 60 hackers took part in the event, most of whom had not previously engaged in the defense space. They formed 14 teams and applied their expertise in artificial intelligence, data mining and machine intelligence to four specific challenges.

Those challenges were to increase situational awareness in the maritime domain, improve the quality of mapping, facilitate better decision making on resource utilization and offer insights to help uphold the armed forces covenant. This is a promise from the nation that personnel who serve or who have served, and their families, are treated fairly.

Getting Down to Business

The hackathon took place over a day and a half and teams built their solutions on the IBM Cloud. Their proposals were submitted to a judging panel that made its evaluations according to criteria set out by the Defense and Security Accelerator (DASA). These included the impact on defense and security, innovation and scientific quality and the likelihood of exploitation.

The overall winner of the most innovative idea was nquiringminds for employing artificial intelligence techniques to analyze Navy Command Automatic Identification System (AIS) datasets to detect anomalies in shipping patterns. Namely, to better identify whether a vessel traveling across global waters is friend or foe.

Typically a patrol boat monitors AIS data from ships within its radar range, looking for strange activity such as unusual geographical position or sudden change in AIS ping frequency. The task consumes a lot of person-hours and is prone to human error.

The winners developed their tool by working on historical AIS datasets. It was able to identify anomalous trajectories based on GPS coordinates, acceleration and AIS ping frequency. The net result was a system that reduces the volume of data that Royal Navy officers have to process and better analyzes the remaining data.

Better Use of Data

At the time of writing this article, there was no word on whether the Royal Navy would make use of the innovation although according to reports the hackathon has boosted the MOD's ambition to make better use of data.

Next Story »