Can a human spot potential shoplifters before the deed is done? Possibly. If given enough training and experience, a person may be able to pick out those who intend to pay for the products and those who don’t. But lets be honest, other than big retail stores, no other retail stores will dedicate staff just to spot and/or deter potential shoplifters. More often than not, this task is delegated to security cameras, but even then, CCTV are there to record the act for the purpose of identifying the perpetrator and possibly, for us as evidence in a prosecution.
However, a Japanese startup, Vaak, has developed an artificial intelligence software that will make the live security camera feeds a lot more useful. Vaak’s software, called Vaakeye, taps into the live CCTV feed to monitor in-store customers to monitor their their body language. According to Vaak, there are telltale signs that signal that if a person may be up to no good. Signs like nervousness, looking around a lot, fidgeting and restlessness could mean the person could be a potential shoplifter and if the analysis by the algorithms turned out positive, so to speak, it will alert the staff on the potential thieves via a smartphone app.
Well, what can I say? This technology certainly sounds like the pre-crime technology straight out of Minority Report (2002, 20th Century Fox), except that it can project images of what’s about to transpire. It merely identify its subject as potential criminal mind. But it don’t matter. The goal here is not to prosecute someone for things they have not done. I mean, it can’t be done. Remember, this ain’t the fantasy pre-crime technology. Instead, it can help to prevent a shop theft from happening. When a staff is alerted to a suspicious character, the staff can quickly move in to approach the potential shoplifter and deal with them like any patron like asking how can the staff be of an assistant. With this interaction, odds are the theft is unlikely to occur.
By combining existing live security camera feeds and Vaak’s software, the incidents of shoplifting could be reduced and with a reported $34 million lost in sales in 2017 to shoplifting, this should be a much welcomed technological development for the retail industry. Vaak is already selling a market-ready version of its shoplifting detection software this month and as of now, the technology has undergone testing in a few dozen stores in Tokyo.
Though, due to it being part of security, it is likely that we will never know if the store we are patronizing is using this technology. But hey, if you do not have ill intents, that shouldn’t be a big of a deal. Well, that’s assuming that the system is 100 percent reliable. That said, next time when a retail staff approaches me in a store, I will think twice about his or her actual motive of asking if he or she could help me. I will be perfectly honest. I’d feel insulted if I was falsely identified as a potential shoplifter. But I guess I will never know, would I?
Featured image and Source: Akio Kon/Bloomberg.