Sheng Siong uses facial recognition to catch shoplifters, identify customers whose payments failed
Sign up now: Get ST's newsletters delivered to your inbox
The system is now in use in around 50 of the 83 Sheng Siong outlets.
ST PHOTO ILLUSTRATION: NG SOR LUAN
Follow topic:
- Sheng Siong has been using facial recognition in about 50 of 83 stores to identify shoplifters and customers with failed transactions since April 2024.
- Number of shoplifters caught doubled from four to eight yearly per store, and failed transaction losses decreased from $10,000 to $1,500 monthly.
- Sheng Siong aims to deter theft and encourage similar technology adoption across retailers, as part of the police's shop theft prevention program.
AI generated
SINGAPORE - Since April 2024, Sheng Siong has added facial recognition technology to its security camera system, to better identify and catch shoplifters.
It is also used to identify unwitting customers whose payment transactions failed, possibly because of network or card swiping issues, before they leave the store.
The system is now in use in around 50 of the 83 Sheng Siong outlets.
Shoplifting suspects are typically identified only after they leave the store with unpaid items.
The staff trawl through security footage and identify the culprit, whose face is then flagged in the system.
When that person visits a Sheng Siong outlet again, the system alerts store managers.
The staff monitor the suspects, approaching and stopping them if they attempt to shoplift again.
Suspects’ faces are removed from the system only after they have been arrested by the police.
Mr Lim Hock Chee, chief executive of Sheng Siong Group, said the system has proven effective.
He said that previously, staff would catch an average of four shoplifters yearly at each outlet.
But with facial recognition, they now catch an average of eight shoplifters at each store every year.
Mr Lim said: “In the past it was a manual process, where we had to print photos of suspects and my staff had to then identify the suspects themselves. Now it’s all auto and digital.”
There were 2,097 cases of shop theft reported in the first half of 2025, an increase of about 4.2 per cent compared with 2,013 cases during the same period in 2024.
Mr Lim said the technology has also helped reduce losses from failed transactions.
He said some customers would unwittingly rush off after tapping their card for payment.
Sometimes, the payments did not go through, but the customers had already left with the items.
Mr Lim said there are about $10,000 worth of such failed transactions every month across all its outlets.
By using facial recognition technology, Sheng Siong has been able to identify such customers.
When these customers return, staff approach them and inform them about the previous failed transaction before requesting they make payment.
This has helped to reduce the amount lost to failed transactions each month to about $1,500.
Sheng Siong intends to roll out the use of facial recognition in all stores.
Posters informing shoppers about the cameras have been put up in all outlets.
Sheng Siong is part of the police’s Shop Theft Awareness for Retailers programme
As at June 30, more than 1,000 retail outlets are part of the programme, including FairPrice, Watsons and Sephora.
On Sheng Siong’s use of facial recognition technology, Senior Assistant Commissioner of Police Devrajan Bala, deputy director of the police’s operations department, said: “Offenders will be identified and caught even if they manage to leave the store with the stolen items. We encourage other major retailers to follow suit on the advancements made by Sheng Siong.”
