Online retailers have been tracking their customers and their web habits with cookies for years. No wonder their brick-and-mortar rivals are looking for new ways to play the big-data game.
The French startup IRLYNX believes it can help them set sales on fire. The company developed small heat sensors, each just 1 centimeter in diameter, that retailers can place on walls, ceilings and even in light fixtures around a store to track customers.
Picking up customers’ body heat, each sensor can monitor movement as far as 15 meters away and within a 120-degree sweep. They can detect heat variances of less than 1 degree Fahrenheit, which helps them tell a human from, say, a hot computer or a fresh cup of coffee.
The sensors can also detect the size and postures of shoppers and distinguish an adult from a child or someone who is sitting down to try on a pair of shoes. The sensors are a big upgrade from the way stores typically track shoppers — with cameras. While the images on a camera may be clearer, it’s very difficult to use those images to track data about how people are using a store. “Video-analysis software can be easily confused by mirrors, photographs, televisions, posters — almost any images of humans,” says Guillaume Crozet, IRLYNX’s vice president for sales and marketing. Training algorithms to disregard these false images can be time-consuming and costly.
Cameras also create privacy issues. The IRLYNX system handles by rendering humans as 16- to 128-square-pixel thermal blobs. The system takes the data coming from sensors and sends it to the cloud, where it can be sorted and analyzed by apps running on Predix, GE’s software platform for the industrial internet. In fact, IRLYNX is one of the startups developing its technology inside the GE Digital Foundry in Paris.
IRLYNX software can combine the feeds from overlapping sensors to create a total picture of the space and how people move through a store over time, Crozet says. The information can help the owner identify how shoppers walk around the store and where they congregate — showing, for example, that people rarely go deeper than the front half of the store, and when they do, they linger in certain areas.
Crozet says that retail is just one of the system’s applications. Earlier this year the company won GE’s Indoor Location Analytics challenge, which provides detailed information on space occupancy. During the challenge, IRLYNX was able to build a proof of concept in its own offices. The company deployed three sensors to track people moving through the entrance and reception area at the Foundry. The sensors collected real-time information about the number of people on the floor and how they were using the space.
The sensors could also help improve the safety of senior citizens living alone and alert caretakers if they detect unusual patterns of movement. The company can even program the system to recognize a distinctive hand gesture, so that users could call for help with a wave of the hand.
Eventually, Crozet says, the sensors can be integrated with GE’s smart LED lights, which use sensors to detect ambient temperature, humidity, carbon monoxide and occupancy, among other things. When connected to the Internet of Things, the sensors can help companies manage their heating and cooling systems, for example. IRLYNX’s sensors take that capacity one step further by tracking people to create a visual record of how people are using industrial spaces. “We want to be at the forefront of digital buildings,” says Crozet. “We want to be installed from the beginning to help companies.”
IRLYNX started working with GE through the company’s Digital Industry Program. You can find more information about the program here.