Gary Harmer of Hikvision UK & Ireland delves into system convergence and how people are utilising open-AI platforms.
There’s no doubt that 2020 was an unusual year, for the most obvious of reasons. But the disruption to normality has not markedly slowed the rate of development in security technology. Current changes are redefining the very scope and remit of the security industry: continuing to protect people and assets but increasingly helping to ensure environments are efficient and intelligent as well as secure.
A key trend is multi-dimensional perception. With the development and application of sensing technologies, more powerful edge computing and intelligent algorithms, we’re now seeing the creation of truly integrated security devices and systems which employ multiple sensors.
Multiple sensors allow additional perception capabilities – such as radar detection, multispectral imagery, humidity and temperature measurement and gas pressure detection – to be integrated with video cameras and systems, effectively extending their detection capabilities and widening the scope of their applications.
Using non-traditional sensors means you can collect and use more information to make more intelligent decisions. Cameras integrated with centimetre and millimetre-wave radar technology can extend perception beyond the visual range to enable object detection and movement tracking up to and over 100 metres away. Other examples include multispectral imagery, for obtaining information in the non-visible light spectrum – using UV detection, for instance, to capture invisible electric arc phenomena, is particularly useful in safety-checks at aging power grids.
User expectations are changing, too. Cameras that capture clear images in all weather and lighting conditions are more likely to be viewed as standard. Low light imaging technology that provides full colour pictures in darker environments and at night is extremely popular.
Additionally, thermal imaging use in extreme conditions like heavy rain, snow, fog or smog continues to be extremely popular. These cameras are far less affected by even the most adverse conditions and ever since thermal imaging has become more affordable, that reliability has been consistently appealing.
The dawn of 5G
5G is here! The much-anticipated wireless standard is being rolled out and it may bring with it some significant changes for the security industry. 5G’s greater bandwidth and lower latency makes the regular transmission of high-quality images possible and, with the widespread adoption of UHD cameras, could bring some new opportunities for video security.
The reliability of wireless transmission via 5G is likely to revolutionise the currently-wired video security market, thanks to the proliferation of wireless cameras and the ability for more edge devices to be connected in remote locations. This will also facilitate the widescale, rapid deployment of AI applications in those edge devices.
Access control systems have long been unwieldy beasts, with the need for cards, tokens and codes creating bottlenecks in environments where security needs to be balanced with high volumes of traffic, such as peak-hour office buildings.
The last couple of years have seen a step-change in the reliability of contact-free facial recognition-based access technology. With faces enrolled on a system, the time taken for recognition and access is minimal and accuracy up above 99%, allowing for almost non-stop entry and exit.
We’ve heard the word convergence for what seems like 15 or more years now. But it’s finally becoming a reality.
Systems working seamlessly together – including video, access control, alarm, fire prevention and emergency management – enhances efficiency and cost-effectiveness. Savings in workforce, installer time, separate maintenance costs, separate software licenses, and so on, all add up to create an attractive package for customers.
Convergence also makes security solutions scale. Existing infrastructure can be united and future needs managed on the same central platform.
Video security is now providing value beyond that of just security. Enterprise-level businesses are being aided in their digital transformation processes and gaining insights into development opportunities.
AI analytics are behind many of these improvements in automation and operational efficiency, in markets including traffic, retail, manufacturing, building, education and more. Retailers, for instance, better understand foot traffic in their stores and optimise their merchandising strategies with help of smart video solutions. Digital dashboards display data and information that comes from integrated in-house enterprise information systems. Users get real-time status updates to help them make the best decisions for their business.
In the cloud
We’ve seen an acceleration in the uptake of cloud-based business activity over the past 12 months, with all sizes of business taking advantage of the cost-effectiveness of cloud services to be more flexible with their operations, deployment and management.
Cloud-based security systems which combine security, networking, storage, analytics and management are making deployment simpler because there’s no need for local servers and software, resulting in time and cost savings.
Cloud-hosting infrastructure also benefits customers with remote operations and maintenance, alerting them to key security events and allowing them to keep up with the latest firmware versions, upgrades and service.
Openness on the edge
With the rise of increasingly powerful edge computing for security cameras, intelligent algorithms are finding wider applications. Intelligent functions like ANPR, automated event alerts, people counting, heat mapping, illegal parking detection and hard hat detection are helping improve security and enhance data system efficiencies.
AI algorithm demand is becoming more diversified and there’s a requirement for application-specific customisation, which has seen an increase in collaboration across the industry. Security manufacturers are increasingly open to third-party AI applications. Customers could help create and train algorithms as well – they have a deeper understanding about their own business, so it makes sense for them to utilise open-AI training platforms to develop their own algorithms based on their data and specific security and business needs.
Cybersecurity and data privacy
Cybersecurity and data privacy protection remains a challenge for the security industry and with the advent of popular cloud-based solutions, the IoT, big data, 5G and AI, millions more security devices are connected every year. As a result, protecting security devices and systems from cyberattacks and establishing data privacy are more important issues than ever.
Cybersecurity will continue to be a concern for the industry throughout every step of data processing, from generation, transmission and storage to data applications and deletion. ‘Zero Trust’ is a popular concept in the cybersecurity field – encouraging consistent verification – and should be embraced by security companies too.