5 MRT stations to use video analytics to detect maskless commuters, unattended luggage

The system will not replace station staff but will instead "be more of an enabler".
The system will not replace station staff but will instead "be more of an enabler".ST PHOTO: MARK CHEONG

SINGAPORE - Unattended luggage and commuters without a mask can soon be automatically detected by a video analytics system used by public transport operator SBS Transit.

The system will be rolled out at five MRT interchange stations - Outram Park, Chinatown, Dhoby Ghaut, Little India and Serangoon - along the North East Line later this year, for a start.

It is expected to be used at interchange stations along the Downtown Line by next year.

The system, which analyses visuals from SBS Transit's existing closed-circuit television (CCTV) network, was developed by the operator and French firm Thales over the last 1½ years.

SBS Transit senior vice-president Jeffrey Sim, who is acting head of rail, said at a media preview last Friday (Sept 3): "Our whole underlying goal is really to enhance customer experience and also to enhance outcomes in safety and security."

The system will not replace station staff but will instead "be more of an enabler", said Mr Sim, who is also SBS Transit's head of the North-East Line/Sengkang-Punggol LRT and head of rail development.

Currently, staff have to patrol MRT stations regularly and monitor CCTV cameras to spot any issues.

The system will help staff identify issues of unattended luggage, maskless commuters and unexpected crowds more quickly.

It can gauge commuter density at the MRT stations - a functionality that was trialled at Woodleigh station in March.

Mr Sim said the system will collect data to identify the usual crowd level at different times of the day. The software will then alert station staff whenever a bigger crowd than usual is detected.

Such an instance could arise if a commuter has fainted.

"The tendency is, when someone has fainted, people will come together and the density will rise," said Mr Sim.

"So that's when the station staff will be able to intervene early, rather than being informed by somebody pushing the panic button or coming up to the station control."

He said the ability to gauge passenger density accurately will help staff decide whether to deploy more marshals or bridging buses to disperse crowds in the case of MRT breakdowns.

Mr Sim said the system is also being adapted to identify commuters with mobility needs, such as wheelchair users or those using crutches. This would let station staff help them more quickly.


The system will help staff identify issues of unattended luggage, maskless commuters and unexpected crowds more quickly. ST PHOTO: MARK CHEONG

Mr Cyrille Bataille, Thales' head of integrated communications and supervision domain in Singapore, said the system can be expanded to add wheelchair identification and other functionalities without requiring additional infrastructure.

SBS Transit is also studying how to make historical and real-time information about commuter density from the new system available to the public, said Mr Sim. This will then help people to better plan their journey to avoid crowds.

Mr Sim said the project was fully funded by SBS Transit, but declined to reveal the cost.

Station manager Roy Ong said the automated alerts by the new system will enable him and his colleagues to reduce time spent monitoring CCTV screens, freeing them to assist commuters.

He added: "I think the mask detection is good, especially with the current Covid-19 situation. It lets us immediately proceed to the passenger and assist them."