Technology could then provide messages designed to promote ‘positive engagement’
Credit: Victoria Jones/PA Wire/PA Images
The government is exploring the use of artificial intelligence technology to detect what proportion of people using public transport are wearing a face covering.
In addition to monitoring use of masks, the tech system could also be used to display messages designed to encourage their use.
According to rail minister Chris Heaton-Harris, work to trial the use of such technology is currently at the stage of being a “proof-of-concept study”.
“[We are] investigating… a non-intrusive AI-model capable of detecting the number of face-coverings, and the number of uncovered faces, in an image,” he said. “The model would then display message responses focused on positive engagement. This work will not be able to identify or track individuals, and no images will be stored by the system.”
Heaton-Harris added that this project – and others exploring the use of emerging technology – are supported by specialist teams at the Department for Transport.
“Technologies such as artificial intelligence and machine learning have many potential applications including in the transport sector,” he said. “Innovation teams across the DfT family support research and development initiatives conducted both within and outside of DfT.”
The minister was answering a written parliamentary question from Scottish National Party MP Chris Stephens, who asked for a breakdown of all AI and machine learning projects currently taking place or being considered by the DfT.
“We don’t hold information centrally regarding further AI or machine learning projects being undertaken or planned by the department at this time,” Heaton-Harris said.
The DfT has previously considered how AI could help alleviate traffic jams, and last year unveiled plans to open up data on planned roadworks for use by external providers of route-finding apps.
Data science professionals at the department have also created a machine-learning tool designed to help policymakers easily identify news stories relevant to their work.