Getting smart about surveillance
Business intelligence is a growth industry with security cameras playing a larger role
By Colin Bodbyl
When most people think of video surveillance systems, they automatically think of security, but surveillance cameras can be used for much more than just security.
The COVID-19 pandemic has spawned innovative new AI that uses the video from surveillance cameras to monitor for compliance with social distancing and mask requirements. While this may appear to be a significant leap for the industry, this type of AI has been around for several years.
Business intelligence or BI data is information collected by technology which is then used to guide the strategy of a given company. An example of this would be a technology that counts the number of vehicles in a big box store parking lot. Daily or monthly changes in the number of vehicles that visit the store could indicate if they are experiencing an increase or decrease in sales. This data could in turn be used to forecast earnings reports which is extremely valuable to financial analysts and investors. With the obvious value of this data, it is no surprise that video surveillance companies have been trying to find a way to use surveillance to collect BI data for several years now.
Simple applications where video surveillance can be used to collect BI data include object counting and heat maps in retail applications. More advanced features include gender or age recognition. In these examples, the data is valuable to marketing teams who may use that information to guide changes to the layout of a store, or the products they sell based on the demographics and behaviours of their customers.
While the value of this data is obvious, it is also a very competitive space. Surveys, loyalty cards, or membership programs collect this same information and with a much higher accuracy level than video surveillance AI. While video surveillance offers a passive method for collecting this data, engaging customer through an active method is not overly complicated and usually renders better results.
In more recent years, video surveillance companies have attempted to collect BI on human behaviour where the people of interest would not otherwise participate in an active solution. One example is tracking the level of compliance with working hours or safety protocols such as what percentage of people are wearing a hardhat or safety vest on a construction site.
In these applications, the value of this data is far more unclear. Knowing that only 80 per cent of workers on a construction site are wearing their protective gear might be concerning, but even if that leads to more oversite which increases that number to 90 per cent, an obvious financial return is difficult to recognize.
Social distancing and mask compliance AI has managed to overcome some of these key challenges. Unlike people counting and demographic data, social distancing and mask compliance is very difficult, if not impossible, to analyze in real-time with other technology. In addition, the potential fines and implications for business owners who do not comply with these regulations are obvious and can quickly justify the investment in a video surveillance system to collect this data.
The business intelligence market is valued at over $20 billion and is growing rapidly. The video surveillance industry seems like it could be a key contributor to that growth with the enormous volume of video data collected each day. However, several leaders in the space have tried and failed over the years to sell business intelligence products at a large scale. The recent pandemic has forced those companies to rethink how we use video surveillance for business intelligence and could just be the boost they need to drive long overdue growth in the space.
Colin Bodbyl is the chief technology officer of Stealth Monitoring (www.stealthmonitoring.com).