Police Scotland is planning to use AI technology with its remotely piloted aircraft systems (RPAS) to find missing and vulnerable people.
The project is being run with a consortium of partners, including not-for-profit innovation centre CENSIS, technology company Thales UK and the University of the West of Scotland.
CENSIS said it is thought to be a first of its kind for UK police forces in using a form of machine learning that provides real time image analysis for identifying humans in rural areas.
It said the core software development work is complete, trials of the new system have begun and the project team expects the technology to be deployed in the near future.
It identifies where a human being is located, rather than an individual, without any facial recognition being involved.
Advance in algorithms
The process has been made possible by new algorithms that can be used on a smartphone or tablet and, unlike similar efforts in the past, does not require computers with heavy processing power.
The software has been trained with hundreds of hours of footage – all taken on police premises – of police officers in different clothing, positions, and situations. It can locate a person within seconds, claimed to be twice as fast as similar algorithms, at a distance of up to 150 metres.
It is also expected that its ability to recognise a human will improve the more it is used.
The RPAS – widely known as a drone – is operated by a specially trained officer on the ground, while another officer receives a real time video feed from the RPAS cameras on a smartphone.
CENSIS said this will help Police Scotland cover large areas of ground in the search for a missing person, reducing the pressure on teams of officers on the ground.
The technology, which is due to be showcased at the CENSIS Technology Summit in Glasgow on 7 November, could potentially be used in a variety of other applications, including monitoring wildlife on land and at sea.
Image from CENSIS