Data storage is at the crux of all surveillance technology
By Jaganathan Chelliah
Looking at the exciting new tech innovations that are changing the world around us, even the past decade seems quaint, almost bordering on the unimaginative. Innovation has also taken the surveillance space by storm. No longer are we using surveillance videos to see what happened, but increasingly deep learning is used to predict what will happen. This ultimately leads to more competent AI that can discern human behavioural patterns.
However, while the concept of analysing video to extract valuable insights is at the heart of deep learning, effective deep learning in surveillance requires both computing power and many hours of “training” video, to analyse shoppers’ behavioural patterns. The insights gained from the captured content in a store or city setting can be applied to everything from retail shopper behaviours, to managing traffic congestion, or providing more efficient operations and yields.
Data is invaluable – New or old
Surveillance systems collect massive amounts of data, but only a fraction of it is utilised. So, what if you need the data for backward-viewing to analyse and understand the context of an incident that occurred? Unless the data is stored, the objective of surveillance is more or less defeated. Additionally, as more and more actionable insights are being extracted from surveillance videos, the importance of saving more and more data is coming to the forefront. As a result, users are holding onto more footage for a longer period.
Right system right storage
These trends reinforce the needs for a surveillance-grade storage that is high capacity, high performance and high endurance to help access and transform the data into valuable insights. Also, while collecting data or video footage is the singular most crucial duty of a surveillance camera, it is also one of the most critical points of failure in many implementations. Relying on the assumed always-on connectivity between the camera and a recording device is one of the biggest risks.
Therefore, always-on-coverage, that is, on-camera storage, in addition to 24by7 back-up storage, is of paramount importance. Today, most cameras come with slot for microSD cards and a few with a slot for SD cards. However, it is extremely crucial that the memory card is specifically designed for recording video surveillance data such as the WD Purple microSD card.
Why do we need surveillance grade memory cards? Because they are designed to deliver superior endurance and reliability, and to withstand extended temperatures, allowing cameras to be placed in very hot or cold environments, inside or out. Importantly, these storage products are built to prioritise write cycles over read cycles, since in surveillance, the focus is on the video being recorded and watched when there has been an incident. So, more write cycles mean more video must be stored, and a non-industrial surveillance grade card is not sufficient.
Help data thrive
From the surveillance perspective, there are many interesting use cases that can provide personalised real-time information where needed, or which provide an added layer of public safety. However, it is important to note that data is the bedrock for these technologies and use cases to flourish. Storage of all that video and metadata allows us the ability to obtain actionable insights from it and to help data thrive.
The writer is director, marketing,
Western Digital India