If a smart city is essentially an intelligent one then only a digitised surveillance framework will help it garner the insights it needs
Deepak Kumar | August 31, 2015
The big cogs in the Digital India juggernaut have started to move. As a great first step, 100 smart cities have been identified. For the number enthusiasts, the five states with the largest number of future smart cities would be Uttar Pradesh at 13, Tamil Nadu (12), Maharashtra (10), Madhya Pradesh (seven) and Karnataka and Gujarat at six each. And yes, as one would already know, each state and union territory would get at least one each.
The telecom and IT industries are enthused – and rightly so. After all, they have been seen as the key enablers of the ‘smart’ cities.
Blueprints are being developed and refined for these projects and a great deal of focus seems to be on making the buildings, work spaces, parking lots, and traffic signalling smart. Attention is also being paid to make use of technologies to improve water management processes, apart from ensuring that cleanliness of public places is duly maintained. And of course, digitisation of the law and order functions is very much an area of focus.
It goes without saying that surveillance frameworks, particularly of the video genre, would be a key to the successful functioning of all these activities and many more. However, video surveillance in its existing form may not be effective enough, especially when it comes to prevention of criminal and terrorist activities. With terrorists becoming more sophisticated in their attacks, the legacy solutions are turning out to be inadequate. While video footage does serve a purpose in helping to nab culprits in the aftermath of an incident, their efficacy in the prevention of the incident itself is little.
Unlocking value of existing surveillance data
A challenge with popular solutions such as closed circuit television (CCTV) is that large teams of trained professionals are also required for a round-the-clock monitoring of the premises where the solution has been deployed. It gets cost-prohibitive as the number of screens and people who monitor those screens grow. While CCTVs are great solution in theory, in practice they have their own sets of limitations.
Is it possible to overcome the existing limitations? How?
A solution lies in the use of video analytics, which works by continuously analysing streams of video data captured by the surveillance cameras. Alerts for events, patterns or behaviours could be triggered as per the monitoring requirements.
Advanced video analytics solutions can be intelligent enough to search for individuals using a set of combined features such as hair, face and torso; or to cause automated alerts if the movement of an object doesn’t conform to predefined patterns.
The analytics engine could be designed to take feed from not just digital IP cameras but also from analogue legacy equipment already installed. This could help modernise an existing surveillance system and enable it with preventive monitoring capabilities. Also, very significantly, the analytics engine could make it possible to search through historical video data and thus vastly improve the efficacy of a repository. In the absence of an intelligent search, it is often very time consuming to sift through hundreds of thousands if not millions of video hours that may have been recorded by a fleet of video cameras over a period of time.
Digitising video surveillance
While analytics would be a primary enabler for digitisation of a surveillance system, a greater value of digitisation would be unlocked once all the five commonly acknowledged enablers of today’s digital systems – social media, mobility, analytics, cloud and sensors (SMACS) – are made part of a wider surveillance framework. How?
Social media: Smart city monitoring agencies could leverage their social media presence to publish city information that is of compelling value to users. In turn, users could also post videos, images or other data they consider to be of interest or importance to a group, community or the city at large. This could potentially turn social media into an extended amateur surveillance arm that works towards common surveillance goals.
Mobility: With the 3G networks nearly ubiquitous in the urban areas and the 4G network rollouts already gaining momentum, smart city projects could benefit in many ways. For example, with a large base of mobile phones equipped with cameras, citizens could capture an incident or any signs of a potential eventuality with their camera phones and post it on social media forums. A greater benefit of mobility could be realised in delivering video feeds and alerts on the mobile devices for the surveillance teams, thus improving their efficiencies and response times.
Cloud: Cloud-based surveillance could exponentially increase the agility of an implementation. Today cloud-based solutions are available that on one hand support video feeds from CCTV and IP cameras as well as other equipment, and on the other hand enable administrators to monitor alerts from any access device such as a PC or a mobile phone. Moreover, advanced analytics applications could be hosted in a cloud data centre and accessed by some or all the smart cities in an as-a-service model. This would also ensure that all cities have access to standardised and updated surveillance software that also meets any regulatory or compliance requirements of central as well as state agencies. Cloud-based implementations could also bring down the cost of storing surveillance data locally.
Sensors: Surveillance has so far largely been about videos and images. However, it could also be a lot about sensors-based data, such as those used for monitoring the utilised capacity of a waste water treatment plant or the water level of a reservoir. Non-video sensors also have a key upside that they generate very little volume of data as compared to video cameras, and as such could easily use even 2G mobile networks to feed data on a continuous basis into the surveillance systems.
Smart cities, on account of their natural adoption of SMACS technologies, would provide the ideal grounds for digitisation of surveillance systems. Moreover, since the smart cities are equitably distributed across all the states in the country, the digitisation would also get contextualised to local and regional needs. Once mature, the solutions would be ready to be ported and implanted into other cities as well.
Smart cities could thus be the catalysts for digitisation of surveillance systems that can be automated to generate near-real time triggers and alerts while also respond to advanced queries and searches seamlessly. At the same time, digitised surveillance could help smart city projects get to speed quicker with regard to their economic, environmental as well as citizen-centric goals. After all, cities that are safe to live, commute, and do business in a sustainable manner only can be truly smart!
(The column appears in the August 16-31, 2015 issue)