When does a camera become an IoT sensor?

When does a camera become an IoT sensor?

When the Internet of Things rears its head in a retail context it’s usually about tracking sensors and RFID tags in the supply chain, but IoT intelligence can also help improve the in-store so-called ‘customer experience’…

Once upon a time, a surveillance camera was a surveillance camera — but (as we know) something changed… and a surveillance camera can now also be an intelligent sensor. So what happens next, how do we populate and educate the intelligence in our new IoT sensors… and are they watching us when we go to the bathroom?

Internet of Business spoke to Prism’s senior VP for global retail Cliff Crosbie who explains, video surveillance cameras used as IoT sensors in retail environments can be used to track how shoppers move around and engage with bricks-and-mortar spaces, all in a way that still protects their privacy.

Prism’s technology itself claims to be able to help us ‘make sense of a visual world’ — the firm’s cloud service turns ‘any’ video camera into a business intelligence tool that can be accessed from ‘any’ device. Retailers large and small, use Prism’s platform to remotely audit, manage, and optimize their real-world business operations.

The use of technologies like this is argued to be giving physical retailers the depth of data previously restricted to e-commerce, along with the traditional sensory and social advantages of good old-fashioned shopping. Surprisingly though, this new technology means customers can shop without fear of ‘big brother’ capturing them on film.

When does a camera become a sensor?

“What most still refer to as ‘CCTV’ cameras have come a long way since the grainy days of analogue. Today’s cameras are mostly digital, very high resolution and now extend beyond just capturing video — to analysing and providing insight into what’s happening on-screen,” said Crosbie.

“Physical retailers are often at a disadvantage to e-commerce sites because they lack tangible data about store activity such as: how many people visit their shops at different times of day and week; where people move throughout shops (pathmaps); where people linger longest (heatmaps), and which items they handle most – all of which Prism can provide.

1slide6One of the most profitable discoveries a retailer can make is which products are getting handled the most, but not purchased. Knowing the underlying reasons – problems with quality, sizing, pricing – a store manager can course correct in time. I’m talking to a large petrol station retailer that has a simple goal of converting £1 of this per customer into sales, which it reckons would seriously boost the bottom line. This is the kind of problem that’s easy to solve with IoT and data analytics technology.

Of course, when retailers improve the store to sell more, the in-store experience always gets better. Popular items that customers want are easier to find, queues don’t build up, and promotions are targeted better. These don’t just directly benefit customers, but they improve staff morale as well, which indirectly benefits customers and the overall brand.” 

Why can’t retailers do this with just cameras?

Prism’s Crosbie explains that, in theory they could use cameras alone, but it would mean having a human being manually pouring over hours of video footage every day and logging all the activity. This would be prohibitively time consuming, expensive, and error-prone.

Other technologies out there can do people counting, heatmaps and pathmaps; however, our technology is set apart by a unique ability to strip out customers’ identities and gather data from their movements. Our approach to privacy means that we adhere to Safe Harbour laws and are the only vendor in our space to penetrate the EU. Our software analyses all the data that the cameras collect and presents it in a simple dashboard format using language and metrics familiar to retailers,” he said.

What if if security needs to ‘get a visual’ on people?

“Herein lies the ‘secret sauce.’ Retailers can set up user access controls to give different levels of access to Prism data. In a typical setup, retailers choose to give the marketing and merchandising teams access to the Privacy Lens, which removes customers from view, but still enables users to render analytics visualisations like heatmaps and pathmaps. However, Security staff in the back office may need access to full video at a certain point in time – say if an alarm was set off. In this case, a security team member could be given access privileges to view Prism through both the Privacy Lens as well as the Enhanced Lens, which does not remove people from view,” he said.

Who is using it then?

Crosbie says that his firm has 300 customers worldwide, including many large, global retailers.

“Unfortunately many big brands using Prism won’t let us talk about it for fear of revealing trade secrets. One I can mention is Londis Stores whose use of Prism led to a footfall increase in stores of 21% — a fifth more than it had previously — and a remarkable 19 percent increase in revenue,” he concludes.