How are edge devices transforming security?

Edge devices are getting smarter as the Internet of Things (IoT) really takes hold. The world is moving to a collection of sensors at the edge that gather data to be processed in Big Data analysis engines, which then derive further benefits from security systems, beyond just augmenting safety and security. Along with safety and protection, we are moving to a world where security data is regularly being used to reduce carbon footprints and improve the utility of the Built Environment. The wealth of information collected by security systems is perfect for this, giving a detailed view of how people and the facilities they use interact, enabling you to optimise their use and the resources you have. Organisations of all types are looking to get real value-add from their security systems purchases and expect to gain better insights into the running of their facilities, through the data they collect.

Edge or not is one of the hottest topics in the security market recently. With the continuous upgrading of the central processing unit (CPU), an edge device will be able to accurately identify and classify hundreds of persons, vehicles, and objects in seconds. It can push real-time events, reports, etc. and achieve complete access control and attendance management for thousands of people. The interconnection of powerful edge devices will also accelerate the security ecosystem‘s development. Customers will centrally manage access control, CCTV, alarms, intercoms, broadcasts, and other security edge devices on one unified management platform. All these advantages will help the customer reduce overall investment. Obviously, for large projects and complex applications, edge devices still face many challenges, such as we have to rely on the existence of back-end management servers. But for the SMB and Residential Security markets, the edge will no longer be a topic, but a trend.

As we increase image processing power at the edge, network cameras can process advanced analytics locally and only send the lightweight results as metadata to the backend server. Without edge-based analytics processing, compressed video streams (H.264/265) must first be decompressed at the server, before server-side analytics can be run. This requires an investment in server architecture and an ability to predict peak demand. Adding more cameras or more analytics requires significantly more server resources until it becomes cost-prohibitive and eventually gates further expansion for many users. Edge-based analytics processing not only saves significant server infrastructure costs, but it also enables a system to scale effortlessly, either on-premise or in the cloud, as an organisation’s needs evolve. As AI-based edge devices become ubiquitous, the ability to proactively alert operators to threats, monitor operational compliance, and supply valuable business intelligence transforms the role of these devices and presents new opportunities for security.

Advances in edge computing have made possible many new applications of security technology. Security devices such as network cameras have become increasingly capable of processing and analysing video at the edge. We are also seeing a new breed of sensors able to analyse sounds, such as gunshots or glass breaks. Also, new industrial IoT devices (temperature, water sensors, etc.) are now connected to physical security systems to go beyond traditional security. With more processing power available on the edge, organisations can easily move towards cloud and hybrid solutions, while running an ecosystem of IoT devices, without burdening their infrastructure. With a non-rigid approach to processing power and a combination of server, cloud, and edge capabilities, organisations can make use of distributed intelligence and operations to increase their security and business efficiency. However, be aware that edge devices can also introduce cyber vulnerabilities, and a resilient cyber security framework has never been more important.

Powerful edge devices are transforming the security marketplace, by making it possible to harvest valuable data in real time. As more cameras are deployed, it’s clear that they can’t all be watched with the same level of focus by human eyes. The latest AI-enabled cameras provide an error-free, proactive, front-line response to threats that help security teams do their jobs more efficiently. By utilising more edge-based processing, we also reduce the need for expensive backend servers. Edge-based analytics can analyse patterns and behaviours that extend beyond security to inform operations and sales teams. This actionable business intelligence might include insights on store layout efficiency and dwell times around product areas. These results then validate assumptions and reveal areas that are not performing as designed. In this way, security cameras have evolved into smart IoT sensors that can share data across departments and contribute directly to bottom line growth.

Edge devices, with respect to video surveillance, usually take the form of cameras with built-in storage for the video data and are managed in a cloud platform. There are some companies making inroads with this technology, but in the way it is currently sold, it is proprietary and expensive. They may claim that they can eliminate the cost of servers, but I believe that is offset by the high prices of the cameras themselves, the licensing costs, and the fact that the video data in effect no longer belongs to the end-user. If you don’t pay your license fee, your system stops working and you cannot access the existing video and, with some vendors, your cameras become bricks.