Earlier in the week Wal-Mart was in the headlines for having a whopping $3 billion dollars in employee theft projected for this year. At first, two things popped into my head: 1) Is that number accurate? and 2) Video analytics.
After giving it some thought, I've decided that the $3 billion is accurate. Given the number of Wal-Mart stores and the pure volume of sales, that number isn't unreasonable. Although, that's still a lot of stealing--and it's coming from the company's own employees.
Thinking about shrinkage (employee theft) and how retailers actually are more often robbed by their own employees than from outsiders reminded me of a conversation I had with Verint's Mariann McDonagh last month when she described some of the applications of integrating video analytics with other systems. One of the most common ways that employees steal from retailers is ringing up a phony "return" when nobody is looking, she said. By integrating video analytics with the point-of-sales system, then the employer can search for all instances of when a "return" was rung up without anyone actually waiting in line. That would be a suspicious situation which could then require further investigation.
This month in Security Dealer I'll have an article that explores some possible directions which video analytics could go. In our industry, it's easy to think about video analytics from a physical security perspective, but that's just the tip of the ice berg. Some people envision video analytics breaking into the mainstream by aiding in the efficient use of resources (both manpower and energy) as well as coordinating better responses to emergency and non-emergency situations.
A few months ago I made a quick stop at a grocery store to pick up a few items before heading to a friend's house. I was in a hurry and knew what I was looking for, so the last thing I was thinking about was the big puddle of water I stepped in right in the middle of the produce section. My foot started to glide and my other leg instinctively lifted as I flailed my arms in an attempt to regain my balance. Fortunately, I didn't wipe out. However, I was so disturbed by what had happened that I resolved that very moment to find an employee and let them know the situation. Interestingly enough, not only was the manager right on the other side of the apples, but he had seen the whole thing in all my awkwardness. (Yeah, I prevented myself from falling but I looked like a clown in the process I'm sure.)
But what if things had been different? What if I had wiped out and been injured--and nobody saw it? Well, some proponents of video analytics see another potential niche for the technology. Theoretically, a camera with video analytics could be programmed to recognize the sudden drop a human makes in a "slip 'n fall" incident, and as such, could alert somebody to the situation. Furthermore, by being able to search for "slip 'n fall" incidents, it could be applicable when the legal issue of liability comes into question.
Although, now that I think of it, I wonder if while stealing a large item an employee has ever slipped and fell? Can you imagine if the video analytics caught that? =)   Â
-Greg
