Singapore’s Patrol Robots: The ‘future-forward’ reality?

The presence of public officers out patrolling on streets to tame and contain public order has never been anything completely unorthodox. Countries have dispensed numerous methods of public surveillance as a means to encourage public safety, from security cameras to social distancing posters plastered around in public spaces.

Singapore has spruiked another innovative addition to its already robust arsenal of  surveillance tools by trialling two robots to assist in patrolling public spaces and counteracting negative social behaviour.

In a statement, Singapore’s Home Team Science and Technology Agency said that they are trialling the pair of bulky autonomously wheeled-vehicles, both named Xavier, to detect bad behavior. These involve breaching COVID-19 safety measures, smoking in restricted sites, illegal hawking, motorised active mobility devices and motorcycles on footpaths or even improper bicycle parking.

Complete with seven cameras mounted on its head and a big tablet lodged in the front, the robots trigger real-time alerts when it identifies “undesirable social behaviour”, which feedback to the command and control centre.

Currently, they are monitoring areas with high foot traffic in central Singapore, a housing estate and a shopping plaza. The robots would then be used for surveillance and presenting messages to inform the public on ‘proper behavior’ during the three-week trial, according to the agency.

The country has earned a reputation for its vast array of surveillance technology, from CCTV cameras to lampposts rigged with facial recognition technology, irrevocably sparking increasing concerns over privacy. 

Officials have taken major strides towards a hyper-efficient, tech-driven “smart nation,” but critics claim that privacy would be compromised and ‘intrusive tech’ would limit people’s control over what happens with their data.

On its recent patrol, one of the robots combed across a housing complex and intervened between a group of elderly people who were watching a chess tournament.

“Please keep one-metre distancing, please keep to five persons per group,” was broadcasted in a robotic voice,  as the camera affixed its lens on them. 

The robots have a variety of sensors installed and are trained on data about the surroundings in order to navigate. 

To determine actions such as smoking or selling goods by screaming, they will film the action and process it with “deep convolutional neural networks and software logics,” matching it to existing data. 

The sample data is collected by engineers walking across the same shopping plaza the robots are patrolling as well as virtually-built synthetic data designed to mimic the actual environment.

Engineers walked the cameras across the commercial mall where the Xavier robots will patrol to acquire some of the sample data. Other training data was created virtually—synthetic data that was supposed to mimic the actual thing.

The devices, according to digital rights activist Lee Yi Ting, are the latest way Singaporeans are being observed.

“It all contributes to the sense people … need to watch what they say and what they do in Singapore to a far greater extent than they would in other countries,” she told Agence France-Presse. But it isn’t a robot cop.

The agency clarified that the robots are not used for law enforcement. Instead, it’s a police aide, a mobile sentry camera and warning system with enforcement powers that involve summoning backup and displaying safety infomercials.

“The deployment of Xavier will support the work of public officers as it will reduce the manpower required for foot patrols and improve operational efficiency,” said an agency spokesperson. 

The government claims that they are necessary to address a labor shortage to compensate for the aging population. 

The true value of Xavier lies in its ability to automate some of the more basic and repetitive policing activities. With human officers in the command center overseeing numerous robots, responses to disorderly behaviour would be better judged from knowing exactly what happened.

Latest