Traditionally, video surveillance was a reactive technology. Aside from a deterrent effect, it didn’t stop incidents happening, but it allowed users to investigate events after they had occurred, often adding evidential footage. When the main security video market transitioned to digital video, the technology became more intelligent, and allowed an active approach to be implemented. As technology has advanced, video surveillance edged towards smarter deployments, but the real change occurred when developments in processing made AI a mainstream reality.
The combination of CPUs and GPUs, coupled with developments in analytics, has changed the surveillance landscape. Today it is possible to create proactive video and audio surveillance solutions which are smart, and exploit the many benefits of artificial intelligence and deep learning. This allows many of the tasks which relied upon operator intervention to be automated, and logical decision-making can assist in the delivery of more secure and user-friendly systems.
Many of the tasks which required users to invest resources (both in terms of labour and costs) can be simplified, and obstacles to the efficient operation of systems have been eliminated. Additionally, systems can now provide a range of benefits for most businesses and organisations, allowing increased efficiencies and better levels of site management. This ensures users can enjoy a return on investment, which isn’t always easy to quantify when a solution is used purely for security.
One option which is aimed at delivering a smart solution comes from Vaion, a new start-up tech company which offers a complete range of smart system elements in its enterprise security solution.
What is it?
Vaion’s enterprise solution brings together a range of elements to create smart surveillance systems capable of delivering intelligent anomaly detection and advanced management features to make event handling a seamless and intuitive task. Combining hardware, software, cloud services and core infrastructure, the solution enables the creation of an end-to-end system.
At the heart of the system is the vcore software, which provides the video and management for the entire system. This can either be installed on the vserver appliance, creating an advanced localised system, or a hybrid solution can be created using both the localised appliances and the vcloud service provision. The final element of the system are the vcam intelligent cameras.
Vaion’s sister company is Jazz Networks, a cybersecurity organisation, and as such the enterprise software has been designed to be hardened against attacks and unauthorised access by default. Where the cloud services are also used, firmware and software updates are also pushed to the system, ensuring there are never issues with out-of-date or potentially insecure applications.
The software makes use of artificial intelligence to deliver benefits such as intelligent incident detection, highlighting of exceptions and anomalies and accurate searches of a wide range of associated data.
Events of interest can be dynamically displayed, ensuring operators are always aware of exceptions or unusual behaviours without a need for the programming of complex rules. Recording is also dynamically controlled, with the system identifying footage of interest (such as motion, exceptional behaviour, incident development, etc.) and allocating increased resources in terms of recording capacity or bandwidth. This allows higher quality real-time streams to be captured when a critical event occurs.
The vserver appliances have been optimised for use with the vcore software, ensuring maximum performance with high levels of reliability and security. The appliances run a hardened Linux kernel and make use of a TPM (trusted platform module) to secure against external threats.
The appliances make use of a CPU and Nvidia GPU combination to ensure adequate processing power for a wide range of analytics-based tasks such as the use of deep learning to identify unusual behaviours.
Initially the cloud service supports set-up, upgrades and licensing, but there are plans to expand this to include remote access and centralised investigation management in the future.
The vcam range includes a 4K UHD static dome and a 360 degree 12 megapixel panoramic camera. The cameras support directional audio analytics, WDR and PoE connectivity. There is a future plan to introduce dynamic encoding, but this is not currently available. The system can be used with third party legacy edge devices if required.
How does it work?
The edge devices make use of integral audio and video analytics to detect threats and generate instant notifications. The analytics are billed as self-learning, so this should reduce configuration time, making the set-up process simpler and faster.
Notifications are dynamic and can either be presented to operator’s via a video wall display, or interactive mapping can be used to highlight incidents. This makes dealing with remote systems a simpler task, as the operator does not need a working knowledge of the site to understand how incidents may be escalating.
Smart recording reduces the overall cost of archiving, ensuring capacity is not eaten up with hours of footage containing little of interest, and live status reports can be shared between a number of operators, delivering more effective protection while also reducing the total cost of ownership of a smart surveillance solution.