Visual data is "conspicuously absent" from many current smart city applications and this could be hampering potential, says James Wickes, Cloudview.
The smart cities concept is beginning to enter mainstream thinking, and it is no wonder. Citizens are starting to experience initiatives that have a positive effect on their daily lives, and city authorities are learning how to derive administrative, financial and ‘liveability’ benefits.
But visual data is conspicuously absent from many current smart city applications, and that hampers how well they perform. Visual data can add layers of detail which help the artificial intelligence that supports many smart city systems to see, understand and interpret with greater accuracy.
Brave new world
Perhaps our reluctance to make use of visual data stems from the notion we have of cameras as ‘command and control’ devices. This is fed in part by the images we see on TV of grainy footage being used to help identify people who are ‘up to no good’, and by reports of visual data being used for purposes other than that for which it has been collected.
Yet what we think of traditionally as the security aspect of visual data and its focus on monitoring individuals is just a small part of what it can contribute. A number of factors are coming together to make it possible for visual data to finally find its feet in the brave new world of smart cities.
The Internet of Things is powering the capabilities of smart cities by allowing information to be gathered from a wide range of sensors, some of them very specialised, combined in cloud-based systems, analysed using artificial intelligence and interpreted ready for automated action or for decision by humans.
Meanwhile data protection legislation which enshrines personal privacy will be strengthened when the General Data Protection Regulation comes into force at the end of May. This places new responsibilities on those who collect visual data to protect individual privacy.
There are many examples of how visual data can support smart city applications in ways that can be of benefit to us all.
The parking problem
According to the British Parking Association, 30 percent of city centre drivers are not on their way to or from a destination but are cruising around looking for a parking space. With cities suffering from high levels of air pollution, and congestion impeding pedestrians, cyclists and other drivers, this is a concern.
So how do drivers find a free roadside parking spot or locate one of the few remaining spaces in a nearly full multi-storey car park?
This is a problem crying out for a city-wide smart solution that incorporates visual data.
Cameras can monitor roadside parking spots, letting a central system know which are unoccupied. Location data can be shared with the driver’s preferred routing app, while visual data can be made accessible so a driver can easily identify the spot they are looking for as they approach it.
Visual data on parking bay use can also help city managers. Rather than sensors just noting that a bay is in use or empty, visual data can identify the type of vehicle – private car, taxi, van, commercial vehicle, disabled driver etc. Using this data in conjunction with a payment app, parking charges can be managed more flexibly, for example varied according to vehicle type, and waived for disabled drivers and some classes of vehicle. Illegal parking can be identified and, with connections to ANPR systems, fines levied. Unused or under-used parking spots can be identified and marketed differently – or repurposed.
Traffic management more generally can also benefit from visual data. For example, those random slowdowns that are caused by sheer weight of traffic can be mitigated if cameras can spot what’s happening and use predictive analytics to ease the flow by changing traffic light phasing.
The analytics can even be weighted to help reduce congestion in what should be pedestrian-friendly areas like shopping streets.
Health and safety
Imagine a cold, wet winter evening. A quiet high street. A lone individual slips and falls. They don’t get up, but there is nobody nearby to assist. Fortunately, a camera has spotted them. The system to which it is connected notes that the person has fallen and is not moving. It sees there is nobody else nearby and decides assistance is needed. It notifies an individual, who verifies the situation and calls first responders. Within minutes our lone pedestrian is being treated by a paramedic for a sprained ankle.
Now think about a large-scale road traffic accident. Visual data of the accident itself as well as the immediate aftermath could be used in near real-time to help first responders decide where their help might be needed most while they are on their way to the scene. Later that same data can be used by accident investigation teams and by city planners who might be able to design out features that could have contributed to the accident.
Health and safety isn’t just about ensuring individuals are cared for at particular times of need. Cameras can observe the general flow of people when there are unusually large numbers in an area, for example at sporting events, and help authorities predictively smooth their paths. On an everyday level, cameras can monitor public areas and automatically notify cleaning contractors when there is a litter problem to be cleared.
The eyes have it
Implementations like these don’t necessarily need a raft of new cameras. Indeed, the existing CCTV cameras that local authorities manage can send their visual data to the cloud, where it can be put to multiple uses across a range of applications, contributing cost efficiency to smart city projects and giving old cameras a new lease of life. The examples above could all be implemented at least partially through existing cameras.
In the old days we thought of cameras primarily as surveillance devices. But as smart city technologies develop and make greater use of artificial intelligence, the whole landscape changes.
Visual data now has the potential to be a key component in the systems that are designed to keep our cities moving, look out for us as individuals, and help us thrive.
So, while it is surprising that city administrations are not yet using visual data as expansively as they could, I think it is just a matter of time before they see the light.