Foodstuffs North Island’s (FSNI) six-month facial recognition (FR) trial has ended with early results showing FR reduced serious harmful incidents by avoiding an estimated 130 incidents such as assault and verbal abuse, leading to improved safety and a decrease in retail crime while respecting the privacy of customers.
FSNI is a co-operative whose members are the Owner Operators of its 320 New World, PAK’nSAVE, Four Square and Gilmours stores throughout the North Island.
In February, FSNI started a trial of FR at 25 New World and PAK’nSAVE stores, to determine if the technology could help stores identify repeat offenders and enhance the safety of both staff and customers.
FR works by matching the faces of people who enter a store against that store’s record of past offenders and accomplices. When the system detects a facial match with at least 90% accuracy, two specially trained team members are alerted and both must confirm a match identification before the information can be acted on.
According to early results from the trial’s independent evaluator, the use of FR avoided an estimated 130 serious incidents such as assault and verbal abuse across the 25 trial stores when compared to previous serious incident data. FR also appears to have deterred repeat offenders, who have subsequently reduced their visits to the stores.
There has also been an 8% quarterly decrease in incidents of retail crime across all FSNI stores, and a 42% quarterly fall in serious incidents.
FSNI Chief Executive, Chris Quin said the early results from the trial showed FR was effective at reducing harm and improving safety but no decisions on its future use will be made until after the co-op receives the final results from the independent evaluator.
“The safety and wellbeing of our team members and customers is our top priority so avoiding an estimated 130 incidents of our people or customers being attacked or abused is a significant achievement,” said Quin.
“We’ve been really concerned about the growing trend of our staff or customers being put in harm’s way. We’ll wait for the final report, but while the numbers are still far too high, it seems like our well-publicised FR trial might have made a real difference.”
During the six-month trial there were 1,747 alerts across the 25 trial stores. For approximately one quarter of all incidents, store teams simply observed the identified match, for another quarter of incidents they took no action at all. In just under half of cases the repeat offender was asked to leave.
Quin said that all too often, it was the same people who kept turning up at a store, despite being trespassed, causing more harm and committing more crime.
“Stores trialling FR became around 40% better at recognising trespassed people compared to our control stores. Information is crucial and knowing an offender is in store means the team are informed and can decide on the safest response, which half the time is simply observing.
“Like any instore security system, FR is as much about deterrence as it is about identification. If it prevents repeat offenders from returning to stores where they’re barred and keeps our people and customers safe, it’s a worthwhile investment.”
FSNI will now await the final report from Scarlatti, an independent research and analytics company, before deciding what happens next.
FSNI General Counsel Julian Benefield said the trial has taught the co-op some important lessons, including that for it to be effective it needs to be supported by robust processes and strong training.
“The trial has been an opportunity for us to learn and improve. Our teams have taken privacy considerations extremely seriously and understand their responsibilities,” Benefield said.
“We’ve proven through our trial that with the right systems, processes and training, FR can help make our stores safer by keeping repeat offenders out. Store teams have told us they feel safer and we know from surveying customers at trial stores that nine out of ten don’t mind it.
“For these reasons, the 25 stores that were part of the trial will continue to use FR in the interim, using the same privacy protocols and processes they used during the trial. The list of stores using FR will continue to be publicly available on our website.”
Benefield said FSNI is waiting for the final report from the evaluators and the outcome of the Office of the Privacy Commissioner’s inquiry into its use of FR.
“We’re respectful of the work of the Office of the Privacy Commissioner and will carefully review their findings. The goal of the trial has always been to establish if FR can help keep our people and customers safe without compromising their privacy. These early results show FR has made a significant difference in reducing harm and improving safety.”
How the trial worked
• Controlled trial
• 25 treatment stores using facial recognition
• Similar number of control stores not using facial recognition
• No sharing of information between stores
• Over 1,747 trespassed offenders successfully identified
