Axis’ Björn Callenfors looks at edge processing and open platforms for creating powerful security solutions including analytics but warns that if an end product seems too good to be true, it usually is.
Physical security has come a long way since its humble beginnings of standalone CCTV systems and manually operated access points. Cloud connectivity, the Internet of Things (IoT) and advancements in the technology itself have transformed physical security into a smart, interconnected system of cameras and sensors. Such systems are now capable of processing data through an analytics engine to produce powerful insights, serving to inform security and operational decision making across many industries and sectors.
Technological advancements to benefit industry
Analytics has many possibilities, some already established and some which have risen to prominence in answer to some of the challenges as a result of the COVID-19 pandemic. The ability to set and monitor occupancy thresholds has proven invaluable in the retail environment as stores re-opened their doors and shop managers needed a way to safely and accurately calculate customer numbers.
In hospitals, mask detection algorithms monitor network video feeds to make sure masks are being worn at all times. Across the public transport industry, analytics can be used to track passenger movement through busy stations and onto their mode of transport, triggering alerts or announcements in response to certain behaviours. These might include a display of aggression, or the breaching of social distancing or face covering guidelines that could impact others negatively in the current climate.
Edge computing makes many new applications of security technology possible. Devices such as network cameras have become increasingly capable of processing and analysing video at the edge. Onboard processing power has huge benefits in relation to the recording and analysis of video in real time and availability of related data. The ability to export and package video surveillance data in the event of an incident, without the potential time lag and energy drain associated with the sending of data back and forward to a server, results in an efficient, cost effective and faster means of video capture, analysis and delivery.
The power of open platforms
In a highly competitive market, those established providers who have realised the benefits of collaboration over a siloed approach are advocating open platforms. An open approach facilitates the sharing of best practices and collaboration over hardware and software. In short, this method enables vendors to come together to deliver a customised best-of-breed solution to meet a customer’s exact requirements. This can be more powerful than anything that one provider could produce in isolation.
Through collaboration and the sharing of ideas, innovation is possible. Much of the current thinking on artificial intelligence (AI) and deep learning in regard to the processing of video and proliferation of powerful analytics comes as a result of knowledge sharing between leaders in these fields. Today’s fluid, dynamic systems can make use of distributed intelligence for maximum levels of security and efficiency, allowing a non-rigid approach to processing power and a combination of server, cloud and edge capabilities. Technology doesn’t stand still and this approach allows developers to make use of a blend of all of these methods of data processing.
AI debunked and the future of analytics
It is important that vendors and end customers do not get caught up in the hype around what analytics will deliver as AI capabilities mature. In order to appeal to a wider market, it’s tempting for vendors to exaggerate the capabilities of their solutions, presenting what is effectively an ‘artist’s impression’ in order to propose a package that seemingly covers all possible bases. Often, such technology is unrealistic in its claims, or represents a serious and oftentimes an indefensible, hardware investment.
In bridging the gap of hype vs reality, it’s important for both end user and vendor to discuss requirements. Vendors should offer a solution that will deliver to time and budget, with full transparency around cost. End users should not be afraid to ask of a provider how far they are away from presenting a fully working version of the solutions they promise. While any invention can be made to sound impressive, it is an innovation, with tangible benefits of scalability, logistics and service, that brings an invention to life as a realistic and deliverable solution.
Smart, secure systems for a safer world
Physical security continues to evolve rapidly, where multiple sensors, intelligence and on-device processing are becoming the norm. With greater emphasis on demonstrating IT maturity and cyber hygiene, building APIs and supporting the latest protocols and languages, the providers of physical security solutions will not only be required to prove their worth in the physical security market, but in the IT space too, as these two worlds increasingly converge. Rather than becoming hung up on specific technology, the security provider of the future will be required to work collaboratively to create a solution to fit a problem, of deploying analytics as a key component and innovating for a smarter, safer world.
Find out more about analytics: https://www.axis.com/en-gb/products/analytics-and-other-applications
By Björn Callenfors, Business Development Manager – Ecosystems, Axis Communications