StopLift is one of those ideas that makes perfect sense and also makes you wonder if you’re being watched. A combination camera/image detection system, it watches self- and staffed checkout systems for signs of theft and “sweethearting” (the process of letting folks walk away with stuff with a wink from the cashier.)
Kundu has a Master’s from MIT and an MBA from Harvard and has been an entrepreneur since graduation. He raised money privately for his project.
“I funded it myself initially, and then raised money from private investors who had made their money in retail and were all too familiar with the deleterious effects of inventory shrinkage which we are addressing,” he said.
The idea came to be when he was reading a Walmart report that said “shrinkage” amounted to $1 billion in annual losses for the chain. He also worked on a “real-time facial recognition system” to look for terrorists in airports. By connecting the dots, he created a number of projects to explore the problem.
“While roughly one third of all inventory shrinkage is caught on video, less than 1% of that video is ever actually looked at. And that is where we saw our opportunity,” said Kundu. “We do hard core video analytics that analyzes the cashier/customers handling of the merchandise while comparing against live POS data.”
In other words, the system compares what it sees with what is sold. As items pass over the scanner, they are checked along with the card and even under the cart. The resulting data is processed in real time and notifies managers when theft happens.
The company currently processes $1 billion in transactions a month and is up against competitors who practice data mining as opposed to direct scanning. It’s a small but fascinating market. The company is working on a way to streamline self-checkout systems using their surveillance techniques, a process that could save retailers time and money. In all, it’s aimed at making your trip to the store more fun and less of a hassle – unless you plan on stealing a ham.