Singapore’s new ‘soldiers’: AI, augmented reality, and data analytics
By Chia Jie Lin
Interviews with the Permanent Secretary of the Ministry of Defence and the Deputy Chief Executive of the Defence Science and Technology Agency (DSTA).
Image: Ministry of Defence/Facebook
For residents of the city, this is difficult to perceive - the city is considered one of the safest in the world. But terrorists networks are becoming ever more sophisticated, Minister for Defence Dr Ng Eng Hen said last month at the inaugural Singapore Defence Technology Summit 2018, organised by DSTA. They use encrypted messenger apps to orchestrate plots, and drones to conduct surveillance and deliver improvised explosive devices.
To combat this, the country is turning to artificial intelligence and augmented reality, and enlisting help from industry to boost digital defence. “We are using AI in facial recognition to deal with terrorists, recognising them,” Neo Kian Hong, the defence ministry’s Permanent Secretary of Defence Development, tells GovInsider on the sidelines of the defence summit.
Scanning for suspicious faces
Singapore is building two platforms that will allow them to track terrorists and differentiate them from innocent bystanders. The police have put up 65,000 surveillance cameras and will install 11,000 more by 2020, equipped with facial recognition sensors that detect and track the faces of suspicious people, the Straits Times reported.
The Government Technology Agency has launched a pilot where over 100,000 lampposts across Singapore will be fit with cameras that link to facial recognition software, supporting anti-terror operations, according to a Reuters report.
Meanwhile, the government’s defence tech arm – DSTA – will use AI to conduct predictive maintenance on navy ships. “AI helps you make better decisions faster,” Hor Gar Yin, DSTA’s Deputy Chief Executive, tells GovInsider.
The agency uses machine learning algorithms to suggest when ships need to be repaired or serviced next, and predict ship failures, according to a DSTA media briefing. It also tracks and predicts how long ships can go without refuelling, and recommend ways to maximise their time at sea.
The system also predicts ship failures by gathering data from previous mishaps, like oil leaks - reducing manpower costs by 30 percent.
New and improved ships
DSTA is currently running virtual scenario-based experiments to build ships that meet users’ needs. In its labs, navy officials can simulate operating a ship even before building it.
Crew members work at booths that simulate their personal workstations. They give suggestions to navy engineers, who then refine the ship’s design for maximum efficiency, significantly cutting ship-building costs. “Augmented reality and data analytics can also help reduce the cost of running the systems,” says Hor. Besides navy ships, DSTA is simulating land and air environments to build better tanks and planes.
DSTA is also using augmented reality to make maintenance work more efficient and safer. The agency’s proof-of-concept app for augmented reality glasses and smart devices provides an animated overlay of the maintenance process, and acts as a step-by-step guide for technicians. The app also provides warning alerts, such as for hot surfaces.
Danger at sea
Singapore is also using data analytics to swiftly identify suspicious vessels sailing near its waters. In collaboration with DSO National Laboratories, DSTA has produced a sense-making system that compares vessel information collected from Singapore government agencies such as the Immigration and Checkpoints Authority and Maritime Port Authority.
The system integrates all these data to identify discrepancies in declared goods, conflicting passports details, and ships that exhibit unusual behaviour.
Working with white hats
Singapore must also build up its digital defences. Here, defence officials must partner with people from outside the government, believes Permanent Secretary Neo.
Earlier this year, the defence ministry invited “white hats” – cybersecurity experts who hack systems to identify weaknesses, he says. The ministry also launched its first-ever bug bounty programme, where people are rewarded for identifying bugs in its systems, according to the Straits Times. “We use the bug bounty programme to invite white hats to test our internet-facing system, so we can find bugs and deal with them,” Neo says.
264 international hackers identified 35 bugs across eight digital systems, including the Ministry of Defence’s website, for a total prize money of S$19,500. Two of these bugs were considered to be of “high” severity.
“All these are expressions of the new way of doing business, of collaborating - inviting more people around the world to test our systems so that we can find where the bugs are and deal with them,” Neo adds.
Sandboxes and experimentation
The defence ministry is also changing the way it is buying technology. “Within sandboxes, we allow certain procurement to be done to test various areas. In all sectors, whether it’s in defence or healthcare, these trials are aimed at deployment. So we should see some of this procurement moving a lot faster,” says Neo.
And the government is collaborating with companies to spur experimentation. “Procuring assumes that something is already on the market, but what if the item is not on the market yet?,” Neo notes. Currently, Singapore is trialling a smart airbase concept, where drones check runways for damage, and recommend taxi routes to cut disruption to aircraft launches.
Meanwhile, a successful trial has in fact led to the implementation of an unmanned smart watch tower at Jurong Island. It guards Singapore’s shores, sending notification alerts to officers in case of intruders. Two more will be set up on the island by September, the Straits Times reported.
Neo adds that the ministry has a sandbox for trialling its autonomous vehicles, and is working with companies such as BMW on this.
With the help of its new digital ‘soldiers’, Singapore can overcome its small size to deter huge threats to national security.