A man, who has been on the run since 2012 for assaulting a police officer with a weapon and drug dealing in Poland, has been apprehended thanks to a new facial recognition app being utilised by South Wales Police.
South Wales and Gwent Police are the first forces in the UK to use the operator initiated facial recognition (OIFR) app, which allows them to identify individuals almost instantly at the touch of a button.
The police app can also be used to identify those who may have given false details, or who refuse to provide details. It can even identify those who are unable to provide details due to vulnerability, unconsciousness or death.
Earlier this month, it emerged that similar software had led to the Metropolitan Police arresting around 1,000 wanted criminals, including paedophiles, rapists and violent robbers.
In South Wales, 50 searches have been conducted using the app, resulting in 10 arrests, two individuals reported for summons, two identified for non-criminal matters and six no further action. In Gwent, the app has been used six times, resulting in three matches. Two were for concerns for safety and one resulted in the arrest of a man who gave false details to officers.
The app correctly identified him as wanted for shoplifting, assault and a motoring offence , reports Wales Online. He was subsequently charged and convicted.
One significant use of the app in South Wales led to the identification of a man in Cardiff who was wanted on a European arrest warrant for drug dealing and assaulting a police officer with a weapon in Poland, dating back to 2012.
The 38-year-old man was remanded in custody to Westminster magistrates' court.
Inspector Ben Gwyer, who spearheads the facial recognition initiative for South Wales and Gwent police forces, commented: "These examples have allowed officers to quickly identify individuals who have refused to give details or given false details before taking appropriate action depending on the circumstances.
"These have included individuals who were found in suspicious circumstances and identified through the use of the app having given false details."
When someone is sought by the police for a criminal offence, the tool enables swift arrest and detention. Mistaken identity incidents are promptly cleared up without the need for a trip to a police station or custody suite.
South Wales Police assures that images captured using the application are not stored and guidance is provided to officers on how to utilise the app in private spaces like homes, schools, medical facilities, and places of worship. This is to strike a balance between policing requirements and individuals' privacy rights.
Inspector Gwyer added: "Police officers have always been able to spot someone who they think is missing or wanted and stop them in the street.
"This technology doesn't replace traditional means of identifying people and officers only use it in instances where it is both necessary and proportionate to do so, with the aim of keeping that particular individual, or the wider public, safe."
South Wales Police says that OIFR is a mobile phone use of facial recognition technology (FRT), which compares a photograph of a person’s face, taken on a police issue mobile phone, to the predetermined image reference database to assist an officer to identify a person for a policing purpose.
Its website states: "Police officers may use the OIFR mobile app to confirm the identity of an unknown person who they suspect is missing, at imminent risk of serious harm or wanted, in circumstances when they’re unable to provide details, refuse to give details or provide false details.
"OIFR can also be used on a person who has died or is unconscious allowing officers to identify them so their family can be traced quickly.
"The OIFR is an overt policing tool. OIFR should be used after an engagement between the Officer and the unknown person has been attempted.
"Independent testing by the National Physical Laboratory found that OIFR achieved 100 per cent accuracy, always returning the correct match as the top result, for persons in the database.
"At the threshold settings we are using, if the person is not in the database, it will not return any results."
