Many airports are currently investing in their video surveillance system capabilities. Common upgrades include the addition of cameras (“streams”), replacing analog or low-resolution IP cameras with 4K or 8K cameras, adoption of IP-based central storage, and the implementation of real-time video analytics.
Depending upon local regulations, some airports have begun to favor long-term retention of video footage instead of auto-erasing it after a set period of time such as 30 days. The increase in camera count, the substantial improvement in camera resolution, and the retention of video is putting a strain on video surveillance systems storage to an extent never before seen. For example, a 4K image typically requires 400% more storage than an HD image. Retaining video for 180 days instead of 30 days increases storage requirements by another 600%. Even compression algorithms are challenged by airport video, given the high degree of motion within airport video streams.
When airports undertake the planning that is needed to upgrade their video surveillance system storage capabilities, they take into consideration many factors. Some of the more common requirements are:
The traditional approach of leveraging tens to hundreds of network video recorders (NVRs) is not feasible as video surveillance requirements grow. Enterprise-class storage architectures are better suited to meet the demanding requirements.
Object storage has emerged as the leading technology for the world’s largest video surveillance systems such as those found in airports, government facilities, and universities, to name but a few.
Object storage creates asset repositories that can hold up to hundreds of petabytes, spreading the data across a large number of disk resources. The data objects can be distributed geographically to provide a self-healing environment that is both more scalable and more resilient than RAID arrays.
Object storage brings several unique capabilities that other storage architectures cannot match, including:
One of the world’s top ten busiest airports, serving tens of millions of passengers per year, initiated a project to upgrade their video surveillance capabilities. This airport had a mix of analog and low-resolution IP cameras. The security management team wanted to upgrade all cameras to be IP based and 4K quality. Most importantly, they had zero tolerance for data loss owing to performance (i.e., the ability to record thousands of simultaneous 4K camera streams), hardware or software issues, or local (physical) incidents.
The airport implemented one of the major video management system (VMS) applications to direct 6,000 simultaneous 4K video streams to “land” on primary disk before it was temporarily cached on Quantum StorNext high-performance shared storage appliances.
Once the 6,000 video streams were temporarily cached on StorNext appliances, the StorNext file system moved the video streams to the 30 PB Quantum ActiveScale object storage system.
ActiveScale’s built-in geospread capabilities (the saving of data as objects that are “spread” among different hardware systems, even different locations) ensured that if a rack of data went offline, no data loss or downtime would be incurred.
Once the video surveillance was stored on ActiveScale, it was accessible by analytics applications that track trends such as foot traffic and facial recognition.
Since the initial phase of the project, the airport has twice added to their ActiveScale capacity in 15 PB increments. This “capacity on demand” approach helped the airport minimize upfront expenditures.In the meantime, 6,000 channels record 4K video 24 hours per day, 7 days per week, 365 days per year with zero downtime. While passengers may not be aware of the forethought that has gone into their safety, it is a compliment to the airport and the security team that make safety their mission.