Data centers are subject to various types of threats. Security, then, becomes a critical element in data center operations. In this regard, video surveillance and AI can play an important role.
Data centers are subject to various types of threats. Security, then, becomes a critical element in data center operations. In this regard, video surveillance and AI can play an important role.
Data centers are places where important and sensitive information is held. As such, they are faced with threats from both physical and virtual worlds.
“Data centers, by virtue of their function, are highly sensitive in nature. The key physical threats to any data center today could be from outsiders coming into the sensitive area, thereby making it imperative to have appropriate security systems to track and authorize people from third-party service agencies. The other more complex threat comes from cyberattacks,” said Abhishek Ghosh, Senior Manager for Marketing and Communications, South Asia and ANZ, at JCI Security Products.
Video surveillance
Indeed, to counter physical threats against data centers, security systems are much needed. As we
reported, perimeter protection solutions serve as the first defense for data centers. Besides that, video surveillance also plays a critical role in data center security.
“It is imperative to have a state-of-the-art video surveillance system, preferably with distributed intelligence, edge-based video analytics, AI-based highly encrypted video surveillance system. This will be used to monitor all public area, entrances, server rooms and racks,” Ghosh said.
“Video surveillance is pivotal in securing data centers from internal and external threats. Today, cameras can perform key analysis on the camera itself and assist in making accurate decisions,” said Anand Chandrashekara, Business Relationship Manager for End Customer, APAC, at Axis Communications.
He added: “Technology advancements have led cameras to be equipped with high-performance processors which assist in processing the data on the edge device and transmit only metadata. Cameras have evolved over the last few years and are available in different form factors to adhere to different needs of customers. They also can produce 4K quality videos and consume less storage space. This enables high-resolution and clear videos for playback and forensic evidence.”
Types of cameras used
Cameras are indeed at the core of video surveillance in data centers.
“Video surveillance cameras can be used for indoor and outdoor surveillance and access for authorized vehicles and personnel. They can be deployed where they are visible or discreet. It also complements solutions like radar technologies and its capabilities to reduce blind spots and false detection alerts,” Chandrashekara said. “Intrusion is one of the leading threats today, and in some data centers, they have a wide perimeter in the location. This could be overcome by having video surveillance across the perimeter, radar, and audio solutions to deter any intrusion.”
For certain areas in the data center, wide area coverage may be needed. This is where
panoramic cameras can come in handy.
“Definitely IP cameras that are minimum 2K resolution (are required). Depending on the customer‘s risk assessments, there can be cameras littered across the aisles and common areas. In those cases, we may mix some 360 degree cameras with regular ones to enhance coverage,” said Patrick Lim, Director of Group Strategy at Ademco Security Group. “A very practical concern is that servicing and maintenance works can be very frequent, and having as complete a coverage can encourage responsible working ethics and accountability, and obviously investigation is much easier with video recordings.”
It’s also advisable to integrate other security system data with video, preferably on a VMS platform.
“At the outset, a robust, strongly encrypted and a completely integrated security solution between access control, video surveillance and to some extent Intrusion alarm detection needs to be in place to protect data centers, preferably on a video management system (VMS) command layer. This will not only help keeping track of all entry and exit points but would also help in tracking and identification of personnel through video analytics,” Ghosh said.
AI
Speaking of video analytics, they are increasingly deployed at data centers as well. A lot of these analytics are AI-based.
“High-quality cameras are absolutely necessary (in data centers), but more than that, video analytics are now a key element in today’s security applications. Innovative digital product and system concepts centered on intelligent video analytics and sensor input allow the best possible coordination of system functionality with operator requirements. This becomes increasingly important for extensive and complex sites with varying security needs,” said Martin Hogberg, Senior Solutions Architect at Siemens.
In particular, AI can work with video surveillance to make sense of video data, helping operators achieve more management and operations efficiency.
“In the past, security industry had been personnel-dependent on identifying and isolating incidents once the incident had occurred. Manufacturers and application developers are now investing in making the process more proactive and assisting decision-making in real time,” Chandrashekara said.
“Quality manufacturers are training their hardware to make their offerings more intuitive and real-time. Adopting AI as a technology by these manufacturers is a natural progression and is welcomed by the industry and customers,” he added. “Data centers are not immune from incidents. The customers welcome any real-time assistance for decision-makers, and they look forward to tech advancements in intrusion, unauthorized access, and incidents.”