How Singapore’s defence agency uses design thinking in AI

By Medha Basu and Ng Yi Ming

Interview with Gayle Chan, Director of the Digital Hub, Defence Science and Technology Agency.

Image: DSTA

Singapore’s Defence Science and Technology Agency is using design thinking principles to build new AI tools for the military, the Director of DSTA’s Digital Hub, Gayle Chan, told GovInsider.

Design thinking is a product design process increasingly used to develop tech products. It involves developing in short sprints and incorporating users’ feedback and experience along the way. “It's just like any agile development type of project: you need the user there; you need to make sure that it's user-centric,” Chan said, speaking on the sidelines of BrainHack, a cybersecurity and AI youth camp by DSTA.

The process of design thinking constitutes going through a user journey, considering key questions such as what are the problems the user wants solved, which of these problems can be fixed through available technology, and how the design team can collaborate with the user throughout the design process. “A fair number of people in DSTA have been trained on that,” she remarks.

Chan also emphasises that the design team itself should be multidisciplinary. This is because the non-tech domain experts are the ones who best understand the areas to which the AI will be applied to. “You need people who are really clued into the business use case, and then you match them up with the right technology and the coders, and with that combined team, then you can get a really good product,” she said.
 

How DSTA is harnessing AI


DSTA is bringing AI into cyberdefence. “In cyber itself, you do need significant amounts of artificial intelligence in order to process the huge amounts of network data that is being generated every day,” Chan explained. Specifically, DSTA is looking to use AI to detect advanced malware.

Beyond the digital realm, DSTA is also harnessing AI, data analytics and Internet of Things to improve fleet management. It predicts failures and picks up defect patterns of defence systems using historical and real-time data. DSTA is exploring how machine learning can be used to detect failures earlier. “The fleet management system is really looking at how we can make our fleet ready - whether it's aircraft, ships or tanks - make them ready at a reasonable cost,” she says.

The third area where DSTA is developing AI is in maritime security. DSTA is looking at using computer vision and machine learning techniques to monitor the waters of the Singapore Strait. “We really feel that there's a lot of applications are on those systems that really run 24/7”, Chan said.

DSTA organised the inaugural BrainHack camp last week to give students an immersive experience of AI, alongside opportunities to get hands-on with coding and pitching competitions. “We wanted to do our part to try and nurture tech talents in this area by creating that environment for kids to grow interest in AI - Singapore needs these tech talents,” said Chan, who was co-chairman of the event committee.

The event brought together over 1,500 students to take part in workshops, competitions and camps on cybersecurity and AI.