Sarah Everard and the eyes that don’t see

Like many women I watched the unfolding of events surrounding the disappearance of Sarah Everard on 3 March with horror. As one of hundreds of thousands of females who often walk to shops, or to meet friends, along typical city and town streets, this was a nightmare scenario that stopped me in my tracks.

But as a female working in the IT software and video surveillance industry, my thoughts quickly turned from shock to hope. After all, with London’s dense network of public and private CCTV systems, from ANPR to private video cameras I believed CCTV footage would be key to finding her very quickly. So did many others I spoke to, even those who usually have ‘big brother’ concerns about CCTV. This was a young woman who just needed to be found safe.

[CCTV in London November 2020[1]]

Except it tragically didn’t happen like that. CCTV did not produce the intelligence required in those first ‘golden’ hours of investigation. In fact, it was less ‘big brother’ and more like ‘long, lost cousin’.

By 8 March the police had to publicly appeal for video footage from the community. Uber even used their smartphone software to pinpoint which drivers had been in the area when Sarah vanished, forwarding dashcam footage they had. Then enormous police manpower was deployed to comb through the hours and hours of submitted CCTV footage to piece together what had happened.

The breakthrough came from a bus camera capturing, by stroke of luck, an image sharp enough to reveal Sarah and the number plate of the rented white car. Too late to save Sarah though.

On a trip to London soon after the arrest had been made, I felt the need to walk the route Sarah had taken, to understand what surveillance was there that should have helped. I noted two public ANPR cameras, four public CCTV cameras, five visible private CCTV cameras and several Ring-type doorbell cameras up to the point where Sarah had last been seen.

One of the public CCTV cameras was for a nearby social housing estate, which I already knew had been found not to be working. The Council did not know it was not working. Two of the private cameras pointing to the entrance and side road of a shop were also hopeless, connected to an old DVR that had recorded over the crucial time period by the time the police enquired.

As I walked along I reflected on the sadness that a tragedy needs to occur to highlight these security failings to their owners, and that it should not happen again.

I contemplated an alternative and more positive scenario where the technology I know exists in the world of video surveillance was monitoring Sarah that night:

  • A world where all the CCTV cameras along the route Sarah took were working perfectly, as they had a) been installed correctly and legally, b) been remotely maintained regularly, c) used software that highlighted to the end users or installers immediately if there were any issues and d) used cloud for event storage so the recordings were secure. There is a feeling of safety in the security being there.
  • I contemplated that the first ANPR camera’s analytics (either alone or combined with artificial intelligence and machine-learning software) had already alerted the CCTV monitoring centre of the hire car being a ‘cruiser’ i.e. continuously driving up and down a particular road or in a particular area (usually the sign of a drug dealer). This would have put the car automatically under surveillance by the system.

The monitoring centre software would then have received alerts every time the hire car was picked up by any of the other public CCTV cameras on the system, meaning a search of all footage where the car appeared could be actioned in seconds.

  • I contemplated the scenario where the police issued an ‘AMBER Alert’ for Sarah. The goal of an AMBER Alert is to instantly galvanize the community to assist in the search for and safe recovery of a missing person, usually a child. These alerts are broadcast through digital road signs, smartphones, and other data-enabled devices.

People with CCTVs along the route Sarah took would willingly log in to their CCTV remotely on their smartphones or laptop, search the timeline the police had given in the appeal, find concrete footage of Sarah within minutes and share it to a dedicated Police incident email with the click of a mouse.

  • The police would apply AI and machine learning to the submitted footage to collate ‘best snapshot’ timelined evidence of Sarah, the perpetrator and the vehicle. High-definition images would be instantly and securely transmitted to neighbouring police authorities and public CCTV monitoring centres to allow their software to find the car or them automatically. They would be discovered.

Today, with Sarah Everard’s killer behind bars, and yet more CCTV being collated around Sabina Nessa’s murder, I feel that the application of existing surveillance technology should be considered more seriously by the public sector to stop the continuation of this awful cycle of events.

But the technology needs to be working.

Just last month I was in Camden reading the local newspaper:

Whether for theft prevention, or our own personal safety, we appear to be lulled into a false sense of security in the presence of physical CCTV cameras when those cameras are not operational. Camden is just one Council that is not aware their CCTV systems don’t work or have ‘legacy’ issues.

The public CCTV is the responsibility of councils and higher priority needs to be given to the safety of public. In particular lone females.

Video surveillance cameras that are monitored, managed and maintained by Tether software become smarter. The whole physical security system (all devices and network) is monitored, alerting users to power outages, internet connection issues or physical security device faults. Tether ensures that ‘best snapshot’ images and footage are captured, stored and available. Tether enables vital evidence is securely and quickly shared with the authorities and integrates with other security devices for a frictionless security service. Tether is a real low-cost solution already in existence that will help councils know whether their CCTV systems are working and compliant.

Security technologies on the cusp of launch include a ‘walk home safe’ app that accompanies you home by monitoring your location and movements using the digital connectivity around you. Another new technology is tethered streetlight drones equipped with cameras and facial recognition software that can find missing persons and identify suspects, respond to an individual’s movement, light the way and guide them to their destination.

I attended an ‘Innovation driving Growth’ event recently where the presenter asked us to think about ‘What is your Moonshot Innovation?’ Not the linear roadmap, but real exponential change, looking at technology that may exist in the near future that you haven’t even thought about. It is estimated that 50% of technology in 10 years’ time has not even been thought about yet. It’s about tackling an issue differently.

It’s difficult to wrap one’s head around ‘Moonshot Innovation’ thinking, but instead of asking the question ‘how can we make women feel safer on the streets’ a question could be ‘how can we eliminate fear forever on the streets’, which could be:

  • Integrated security devices everywhere eliminating crime
  • Floating safety bubbles protecting lone walkers
  • Patrolling Robocops
  • Eliminating the need for streets
  • Time travel or virtual bodies, negating the requirement to be on streets

We need a new operating model for exponential change, so I for one will leave the ‘Moonshot Innovation’ to others. But for now we owe it to the Sarah's and the Nessa's of this world to apply the security technology we know exists to the surveillance infrastructure we have to make it safer for all.

The eyes need to see.

[1] November 2020