May 15, 2023
For supermarkets, past few years have been challenging: first, online shopping has taken away many young shoppers; space is gradually shrinking; thirdly, losses of supermarket are huge, especially act of theft, which places an additional burden on supermarket. The third point is most personal pain of every retailer. Because it's an unnecessary expense, but it's an unexpected huge expense.
According to 2014-2015 Global Retail Theft Barometer Annual Report, shoplifting is leading cause of losses in China's retail industry. China's losses from 2013 to 2015 amounted to 50.447 billion US dollars. In recent years, there has been an upward trend from year to year.
Generally speaking, in order to reduce number of food thefts and avoid losses caused by thefts, many supermarkets usually set up a loss prevention department to catch thieves. This is not only common sense, but also a manifestation of right to self-help. But it is worth noting that some supermarkets deal with thieves privately and impose high fines who are suspected of breaking law or even committing crimes and should be abandoned.
Besides loss prevention, electronic anti-theft systems are most popular measure, with 73% of supermarkets using them. In addition, inventory management technology and radio frequency identification (RFID) etc. are also means to prevent theft in supermarkets. However, these common practices have not really prevented theft, on contrary, law is a foot above devil, and supermarkets urgently need new safeguards against theft.
An AI+Big Data model can predict occurrence of theft, and Japan already has practice using this model. NTT has partnered with Earth Eyes to create an "AI Guardman System" that captures all customer movements with cameras and correlates them with "suspicious" behavior to track suspected thefts in real time. The system helped stores reduce shoplifting by about 40% during testing phase.
Artificial intelligence and big data help supermarkets
1. Automatically take photos of faces of customers entering store, compare and identify them;
2. Suspicious personnel alert, alert information is sent to mobile phone in real time;
3. Detect suspicious persons in a timely manner, focus attention on them and prevent them in advance;
4. Help stores catch criminals and cut losses;
5. Increase security and reduce burden on employees;
6. Save trace evidence to make future traces easier;
7. Keep a record of information that is convenient for next warning and prevention, as well as reprisals against repeat offenders.
Because China is very tolerant of facial recognition technology, 83% of retailers in China are ready to participate in creation of facial recognition databases. However, according to current technology, these measures cannot effectively eliminate external interference, it is often impossible to recognize products or faces in crowded supermarkets, so level of this technology still needs to be improved.
With regard to demand for anti-theft products and people's deep understanding of high technology, it is believed that security measures more suitable for anti-theft protection in supermarkets will be gradually put into practice. Perhaps in near future this problem can be fundamentally solved thanks to technology.