OPINION James Wickes, CEO and co-founder of Cloudview, explains why visual data is an untapped resource for smart analytics within many IoT projects.
An occasional series of vendor perspectives on the world of connected business – because it’s all about making new connections and starting new conversations.
We are constantly reading about IoT developments, but these rarely include visual data – which is strange, because sight is our most powerful sense and we are surrounded by digital cameras. However, much of the visual data currently collected is stored locally and only used for a single purpose, while a huge percentage is never used at all. Combining this with other IoT data streams and adding analytics would make it immensely valuable.
The volumes of visual data available are eye-watering. Looking at CCTV alone. In 2015, the British Security Industry Association estimated that there were between four and six million security cameras in the UK. Our own research suggests there are now around 8.2 million. Even six million cameras recording 12 hours a day would capture 72 million hours of footage every day, producing 7.5 petabytes of visual data every hour.
Analytics and visual data: a formidable pairing
Applying analytics to visual data is complex. However, we now have the processing power, bandwidth, data storage capacity, and computing ability to enable fast, reliable analysis to a standard that makes it commercially viable. McKinsey expects video analytics to experience a compound annual growth rate (CAGR) of over 50 percent over the next five years.
Adding analytics and cloud storage to cameras provides the ability to spot anomalies that we are unable to identify with our own eyes. For example, in health and well-being alone there are many opportunities, such as:
• A camera trained on a patient in a hospital with the right analytics can now spot irregular breathing or an irregular pulse.
• Cameras are being used in care situations to monitor individuals to ensure they are being well-treated (with appropriate permissions).
• Qualified health and social care professionals are able to review footage for safeguarding purposes, and this can prove popular with both residents and staff.
Building the VIoT
The next step is to combine visual data with other data sets – from static data, such as grid references, to dynamic data, such as weather information.
This will create a vast new market – the Visual IoT (VIoT). In other words, the integration of visual data into a uniform, IP-based data stream, combined with the capabilities and functions of a network of physical objects and devices.
In this way, cameras can be turned into super-charged sensors providing data that can then be acted upon, such as identifying that a car with a certain numberplate is allowed to enter a given area, which automatically opens the gate.
The potential is huge, and could revolutionise traffic management, and the reporting of crimes or accidents. For example, when an individual with a VIoT device enters a certain area, by previous agreement their data could be aggregated with that of others to create an accurate picture of an event.
For a motorway accident, combining data from road cameras and in-vehicle routing systems would pinpoint the precise location and help first responders to arrive more quickly. Meanwhile, adding visual data from drivers’ dashcams (with permission) could add unique views of the area around an incident.
Combining visual data with analytics can provide insight into both what is happening and why things happen, together with the ability to anticipate what might happen next.
Consider the control centres used by emergency services to monitor cameras in city centres. Adding analytics and machine intelligence would enable them to identify impending problems and send resources to defuse a situation before it escalates. The same process could identify potential risky or suspicious behaviour at transport hubs and other public spaces.
There is also tremendous potential for smart city initiatives that use existing camera data to improve the local environment. For example, NVIDIA is developing an intelligent video analytics platform for smart cities, which will apply deep learning techniques to video streams. Applications include public safety, traffic management, and resource optimisation.
Safeguarding privacy and GDPR
The big issue, of course, is privacy, but technologies such as facial and behaviour recognition can be used to reduce human involvement to a minimum. The General Data Protection Regulation (GDPR) provides additional protection, as it includes provisions for how visual data is collated and used in applications that apply AI, analytics, and deep learning techniques to that data. There are also applications in sectors such as the environment that will not involve individuals at all.
Provisions such as privacy by design, Privacy Impact Assessments, and the appointment of a data protection officer will be mandatory for public authorities and any organisation whose core activities require regular and systematic monitoring of data subjects on a large scale. There are also applications in sectors such as the environment that will not involve individuals at all.
By providing information that is not available in any other way, visual data will enable the IoT to bring even more benefits to all our lives. More information is available in the white paper Visual IoT: where the IoT, cloud and big data come together.
Internet of Business says: This opinion piece and the link to an external white paper have both been provided by Cloudview, and not by our independent editorial team.