printer friendly version

Are we over the edge?

The surveillance market has seen some dramatic advances over the past few years, offering users more options and better services from their solutions. Far from putting video on a screen or saving it to a storage device, surveillance technology delivers many functions today. From visual verification or even facial biometrics on a camera to advanced analytics and AI enhancements, and even going beyond security to offer operational features, the capacity of the old CCTV camera has advanced tremendously.

One feature that has received attention over the years is the ability to do more in the camera. Initially, it was on-camera storage, where users could store high-resolution video on the camera, only accessing it if needed, while a lower-resolution stream or perhaps individual images were sent to a control room. Analytics on the camera enhanced this ability, saving bandwidth by storing video on the camera and only alerting operators when the software picked up an anomaly.

Increased processing power made it possible to run AI on cameras for more reliable analytics. Azena even offered a camera operating system and an Application Store where companies could sell AI analytics as apps and users could use them as and when needed (on cameras with the organisation’s Azena Open Camera Platform). The store is still available at https://store.azena.com/, although the drive to establish the ‘open OS’ seems to have lost steam and been incorporated into Bosch Security (support is still maintained).

With the advance of cloud services, the surveillance industry was yet again changed. Cloud has overshadowed edge processing and storage although it is taking longer to become standard in some African countries. Yet, users are no longer buying on technical specifications or potential functionality but are more interested in broader solutions that solve their problems. This has not dampened the drive to the cloud but has ensured that edge devices are used more sparingly where the use case fits.

To find out more about edge surveillance, and not forgetting that many users still want their surveillance onsite and under their control, no matter how much bandwidth they may have or how reliable cloud is becoming, SMART Security Solutions asked Axis Communications and Hikvision for some insights. We wanted to know what users are looking for, how they approach cloud/edge/onsite options, what the manufacturers provide, and how they deal with hybrid approaches. This article focuses on edge surveillance, but a solutions-oriented market requires that we include everything.

Hybrid is complicated, but efficient

Axis Communications’ Rudie Opperman says the surveillance market is evolving rapidly, with organisations weighing the benefits of cloud, edge, and onsite solutions. “While cloud adoption is increasing, the hybrid approach remains the most practical solution for many end users. The growing demand for real-time analytics, cybersecurity and cost-effective storage drives continued investment in edge solutions.”

He adds that cameras with built-in edge processing – leveraging AI and analytics – are widely deployed in sectors where real-time decision-making is critical, such as safe/smart cities, transportation and critical infrastructure. These deployments allow organisations to filter and process video at the source, reducing bandwidth usage and ensuring faster response times.

Hikvision’s Ruaal Tromp notes, “Like AI, the cloud is not a new trend in our industry but an expanding one. From small business markets to the enterprise level we can see that more and more businesses are leveraging cloud-based security solutions and services.”

However, deployment depends on specific needs. The market is evolving toward a balanced adoption of edge, cloud, and hybrid solutions. While cloud-based systems offer scalability and remote accessibility, edge computing addresses latency and bandwidth challenges by processing analytics locally. Hybrid models are gaining traction, particularly in sectors like retail that require real-time analytics at the edge combined with cloud storage for remote management and long-term insight analysis.

The storage and bandwidth question

While the market has thankfully passed the dial-up and GPRS stage of bandwidth availability there are still limitations, especially when considering the quality of collected video (which takes up space on storage and in transit), multiplied by the number of cameras. A pure cloud approach will require significant reliable bandwidth with back-up options, which makes edge and onsite systems more attractive.

While cloud storage offers scalability, Opperman agrees that the associated bandwidth requirements for full-resolution video streaming present challenges. “Many end users prefer to store critical footage onsite or at the edge while leveraging cloud platforms for analytics and long-term archiving.”

Edge storage ensures data remains available even in cases of network failure, enhancing operational resilience. Full control-room operations in the cloud are theoretically possible, but they are not yet practical for many applications due to latency concerns and bandwidth constraints. “A balanced approach that combines edge storage with cloud-based analytics provides the best of both worlds,” says Opperman.

Tromp also notes that storage preferences depend on specific use cases and are restrained by storage and bandwidth conditions. He agrees that “while fully cloud-based control rooms are technically possible, most operations still maintain local infrastructure for mission-critical functions.”

Distributing AI to the benefit of the user

No discussion of surveillance (or almost any industry today) can occur without looking at the impact of artificial intelligence and how users can obtain the best AI performance for their money. Numerous companies offer AI analytics in the cloud, often relying on images and small snippets of video instead of a full HD feed. As noted above, edge AI offers immediate analysis and enables faster response. So, which option is a better choice?

Tromp explains that Hikvision provides converged cloud-edge solutions that enable faster, more efficient services. This approach delivers real-time intelligent capabilities, from smarter perimeter protection to cloud-based security management, while enabling immediate analytics and better edge decision-making. “Our cloud platforms additionally reduce hardware requirements, offering scalable solutions that lower both initial and operational costs for businesses of all sizes.”

“At Axis Communications, we recognise the need for scalable AI-driven surveillance solutions,” says Opperman. “Our cameras incorporate advanced edge analytics in two ways: free and bundled, as in the case of Axis Object Analytics. We also enable technology partners to port more advanced licenced edge-based analytics onto our devices.

“In both cases, video data is processed at the source to reduce bandwidth strain while enabling real-time insights. These devices can work in tandem with cloud-based AI for deeper analysis, pattern recognition, and long-term trend analysis. This distributed intelligence model ensures flexibility, allowing organisations to maximise efficiency without solely relying on cloud infrastructure.”

Updates and the app approach

A benefit of onsite or cloud servers is that the latest analytics and AI algorithms can be installed and updated as required, or automatically in the case of cloud services. In the Azena approach, edge devices could be easily updated, and an app could be swapped out for a different app. Azena has not gained the desired traction, perhaps because many leading camera manufacturers did not support it, but the idea is still alive.

Tromp says this flexibility is supported in some Hikvision models, such as the latest Dedicated DeepinView Cameras. “Some models feature six switchable deep learning algorithms in a single unit. Users can manually enable or switch algorithms as needed, for instance, changing from people counting to perimeter protection when redeploying a camera.”

For Axis, flexibility and scalability are key as edge solutions need to evolve with user needs. Opperman adds that Axis edge devices support an open platform approach called ACAP (Axis Camera Application Platform) allowing customers to install and update AI applications as needed. Users can update the analytics accordingly if a camera needs to transition from people counting to licence plate recognition (LPR) due to a change in its role. This adaptability ensures long-term value, reducing the need for hardware replacements while enhancing operational flexibility.

“We offer secure and flexible device management tools to perform this change of analytics, as is the case with Axis OS updates and other changes to ensure easy deployment. There are, of course, huge advances in hardware capability that could also add tremendous benefits, such as the device simply needing to be replaced with the latest chipset, but often, the older devices are deployed somewhere else where they can still add value.”

What’s in the box?

Our final question to both companies concerned the products and solutions each offers for edge surveillance. Moreover, we asked whether edge devices are compatible with third-party platforms or are they limited to the company’s systems.

According to Opperman, Axis Communications offers a comprehensive portfolio of edge-enabled surveillance solutions, including cameras with built-in AI, edge storage, and cybersecurity enhancements. “Our devices are designed for seamless integration with various video management systems (VMS) and cloud platforms, ensuring compatibility across different ecosystems.

“The recent launch of the ARTPEC 9 system-on-chip (SoC) supports cloud-friendly AV1 video compression, yet still supports H.264 and H.265 for compatibility with older VMS. This open architecture enables customers to build surveillance infrastructures that best fit their unique requirements without being locked into proprietary solutions.”

“We provide complete AI-powered solutions, including AcuSense technology, the DeepinView cameras and the DeepinMind servers,” says Tromp. “These solutions automate security measures and business operations through features like smart alerts, false alarm reduction, and people counting. Our edge devices maintain broad compatibility while offering deep integration with Hikvision’s management platforms for optimised performance.”

While this article focused on edge surveillance, Opperman notes, “the future of surveillance is not a question of edge vs. cloud, it’s about how these technologies complement each other. Edge devices remain highly relevant, providing real-time intelligence, bandwidth efficiency and enhanced cybersecurity. At the same time, cloud services offer scalability and long-term analytics capabilities. A hybrid approach ensures organisations can leverage the best of both worlds, creating a surveillance ecosystem that is secure, flexible, and scalable.”

For more information contact:

• Axis Communications, +27 11 548 6780, [email protected], www.axis.com

• Hikvision South Africa, +27 87 701 8113, [email protected], www.hikvision.com

Share this article:

Further reading: