The term “smart city” gets thrown around a lot nowadays, but as different technologies that strive to be defined in this way are adopted by different countries globally, the meaning of this phrase gets lost in translation. The simplest way to define a “smart city” is that it is an urban area that uses different types of data collecting sensors to manage assets and resources efficiently. One of the most obvious types of “data collecting sensor” is the video camera, whether that camera is part of a city’s existing CCTV infrastructure, a camera in a shopping centre or even a police car’s dash camera.
The information gathered by video cameras can be used with two purposes in mind, firstly: making people’s lives more efficient, for example by managing traffic, and secondly (and arguably more importantly): making people’s lives safer.
In the smart and safe city, traditional record-only video cameras are of limited use. Yes, they can be used to collect video which can be used for evidence after a crime has taken place, but there is no way that this technology could help divert cars away from an accident to avoid traffic building up, or prevent a crime from taking place in the first place.
However, streaming live video from a camera that isn’t connected to an infrastructure via costly fibre optic cabling has proven challenging for security professionals, law enforcement and city planners alike. This is because it isn’t viable to transmit video reliably over cellular networks, in contrast to simply receiving it.
Transmitting video normally results in freezing and buffering issues which can hinder efforts to fight crime and enable flow within a city, as these services require real-time, zero latency video without delays. Therefore, special technology is required that copes with poor and varying bandwidths to allow a real-time view of any scene where cameras are present to support immediate decision making and smart city processes.
There are many approaches to transmitting video over cellular. A specialist codec (encoding and decoding algorithm) has been developed that can provide secure and reliable video over ultra-low bandwidths and can therefore cope when networks become constrained. Another technique, which is particularly useful if streaming video from police body worn cameras or dash cams that move around, is to create a local wireless “bubble” at the scene, using Wi-Fi or mesh radio systems to provide local high-bandwidth communications that can communicate with a central location via cellular or even satellite communications.
Live video streaming within the smart and safe city’s infrastructure means that video’s capabilities can go beyond simple evidence recording and evolve into a tool that allows operations teams to monitor and remediate against incidents as they are happening.
This can be taken one step further with the deployment of facial recognition via live streaming video. Facial recognition technology can be added on to any video surveillance camera that is recording at a high enough quality to identify faces. The technology works by capturing video, streaming the live video back to a control centre and matching faces against any watch lists that the control centre owns. Importantly, the data of people who aren’t on watch lists is not stored by the technology.
This technology can work to make the city safer in a number of ways. For example, facial recognition could spot a known drug dealer in a city centre where they weren’t supposed to be, or facial recognition could identify if a group of known terror suspects were visiting the same location at the same time, and this would send an alert to the police.
In an ideal world where the police had an automated, electronic workflow, the police officer nearest to the location of the incident would be identified by GPS and would be told by the control room where to go and what to do. Most police forces aren’t quite at this technological level yet, and would probably rely on communicating via radio in order to send the nearest response team to the scene. As well as this, shopping centres could create a database from analogue records of known shoplifters to identify criminals as soon as they entered the building. This would be even more effective if run co-operatively between all shopping centres and local businesses in an area, and would not only catch any known shoplifters acting suspiciously, but would act as a deterrent to shoplifting in the first place.
As mentioned above, live streaming video from CCTV cameras can help the police fight crime more proactively rather than reactively. This can be enhanced even further if combined with live streaming video from police car dash cams and police body worn cameras. If video was streamed from all of these sources to a central HQ, such as a police operations centre, the force would be able to have full situational awareness throughout an incident. This would mean that, if need be, officers could be advised on the best course of action, and additional police or other emergency services could be deployed instantly if needed.
Incorporated with facial recognition, this would also mean that police could instantly identify if they were dealing with known criminals or terrorists. Whilst they would still have to confirm the identity of the person with questioning or by checking their identification, this is still more streamlined than describing what a person looks like over a radio and then ops trying to manually identify if the person is on a watch list.
The smart, safe city is possible today – for one, if live video streaming capabilities are deployed they can enable new levels of flow in the city. With the addition of facial recognition, cities will be safer than ever before and law enforcement and security teams will be able to proactively stop crime before it happens by deterring criminal activity from taking place at all.
