Inside Edinburgh’s real-time operations centre

Edinburgh has opened a new City Operations Centre that brings in real-time CCTV data with video analytics for public safety, transport management and urban planning.

The initiative is part of the Scottish capital’s smart city strategy.

Council Leader Cammy Day told Cities Today: “I think it’s one of the leading projects in local government.” He said it will enable the city to better manage large events such as Hogmanay and the August festivals, optimise transport and active travel infrastructure, and even help to find missing people faster – such as children or those with dementia.

According to Day, the round-the-clock facility could help operators to “identify problems before they even happen”.

The launch of the centre follows a recent CCTV upgrade, which replaced analogue cameras across the city with high-resolution, internet-enabled devices and introduced new video analytics capabilities.

“This is not a CCTV system anymore. It’s much, much greater than that,” said Day, adding that it will provide more data than the city has had before.

Beyond public safety

Miranda Matoshi, Edinburgh’s European Regional Development Fund Project Manager and lead on the CCTV modernisation programme, described the operations centre as a “central nervous system” and said it was important to use the upgrade to “leapfrog” ahead rather than just replacing infrastructure.

“It was about really trying to make the investment work harder for us,” she said. “While CCTV has historically been focused on antisocial behaviour and crime, and that’s still its primary purpose, the clarity in the images means that you can then run video analytics software, across the recorded and live footage, to help the operators actually view 200 CCTV cameras throughout the city, which would be impossible to do manually all the time.”

This could include detecting issues such as crowds gathering or a road collision for faster response.

Operators could also, for instance, search footage for a red car or a child in a blue top to speed things up in critical situations.

Matoshi believes this functionality could reduce manual review time by “a factor of ten”.

“When it comes to video analytics, the sky’s really the limit and it’s about how you then calibrate it to deliver the insights that your city specifically requires,” she said.

Plans are underway to generate heat maps which show different modes of transport, such as walking, cycling and cars, to support planning decisions.

“I am hoping that we’re going to be generating datasets that will be able to help colleagues across the organisation,” commented Matoshi.

Currently, the city is only running analytics on one camera but plans to increase this to 100 by mid-2023, with scope to scale as needed.

The system also gives the city more autonomy. Matoshi said vendors often offer to install cameras, run analytics for the city and sell them back the data.

“Ultimately, if we’ve already got the CCTV network, why are we not doing that in-house and leveraging the infrastructure we already have?”

Privacy

The £2.6 million (US$3.2 million) programme was delivered in partnership with IT service provider North, funded by a £712,000 grant from the European Regional Development Fund (ERDF) and the rest from the council.

There are around 200 permanent public realm cameras, including pan-tilt zooms (PTZ), and an additional 100 static cameras at existing locations.

Matoshi said the static cameras are important because they capture consistent data for analytics.

“I think that’s perhaps a hurdle that some local authorities have fallen at before where they’ve installed lots of new pan-tilt-zoom cameras,” she said. “As soon as you move the camera, you’re breaking that consistency and then you’ve almost stopped that dataset.”

The city stresses that the system is compliant with GDPR and other relevant legislation, and said that facial recognition and automatic number plate recognition are not being used. Still, some civil liberties organisations warn that video analytics represent a fundamental shift in the nature of CCTV.

“Cameras that collect and store video just in case it is needed are being transformed into robot guards that actively and constantly watch people,” the American Civil Liberties Union stated in a 2019 report titled The Dawn of Robot Surveillance.

Matoshi said she is proud of the project’s privacy and security by design approach, including a data protection impact assessment on the video analytics specifically.

“There are no autonomous steps involved,” she said. “It’s just there to support our operators and what they already do.”

The city is also exploring how to integrate the data with information from housing and waste sensors which are being rolled out to detect issues such as the need for property repairs or bin collections.

In addition, the project will make datasets such as pedestrian, cyclist and vehicle counts publicly available.