Marcelo Ang Jr.
Department of Mechanical Engineering
National University of Singapore
AsianScientist (Feb. 28, 2020) – Remember the Jetsons? Television audiences first encountered the futuristic family at the height of the Space Race in the early 1960s. Nearly six decades later, our visions of the future still mirror the Jetsons’ candy-colored utopia—except today, we’re intriguingly close to turning their fictional technologies into reality.
Though flying cars and robot housekeepers like the Jetsons’ Rosie still remain out of our grasp, robotics researchers like Associate Professor Marcelo Ang Jr of the National University of Singapore (NUS) are diligently advancing the field, one prototype at a time. In 2015, Ang and his team launched one of Singapore’s first trials for driverless cars, also known as autonomous vehicles (AVs). The following year, his team unveiled a self-driving scooter that is now ready to go on the market.
But Ang has big plans for robots beyond the road. As Director of NUS’ Advanced Robotics Centre (ARC), he leads research on human-centered collaborative robotics, in hopes of someday making seamlessly symbiotic human-robot interactions possible. A passionate advocate for amplifying public interest in robots, Ang is also the founding chairman of the Singapore Robotic Games—a yearly Olympics-style tourney where robots run marathons, climb walls and even sumo wrestle.
With robots slowly but surely moving into the mainstream, Ang candidly shares with Asian Scientist Magazine his views on the wider relevance of robots in contemporary society and gives us an inside look at his upcoming projects.
- What sparked your initial interest in robotics?
- Though robotics encompasses a wide variety of applications—from humanoid robots to automated assembly machines—you’ve become known for developing a variety of autonomous transportation systems. What inspired you to specialize in AV research?
- Unfortunately, Asian megacities tend to dominate lists of the world’s most traffic-choked cities. How can AVs help ease congestion in these cities?
- What remaining challenges do AVs need to overcome before we see their widespread adoption? When do you think they will become mainstream?
- What are some pressing ethical issues surrounding robots?
- Aside from AVs, your interests also span automation and AI computer control. Could you share with us other interesting projects are you working on right now?
- What other innovations in robotics and AI are you most excited about?
- Emerging technologies such as robots, AVs and AI may be intimidating especially in settings where they aren’t as common. How can we make such concepts less intimidating, especially for aspiring scientists?
I’ve always been interested in computers, but they just move data around. So I thought, what if data can also be used to physically move machines? After graduating from De La Salle University, I worked at Intel Manila and had one of the first IBM PCs in the 1980s. There, I was exposed to the latest computer tools.
Then I went abroad. Though my background is in mechanical engineering, I did a PhD in electrical engineering, because robotics marries mechanical, electrical and computer science. In 1989, I was recruited to Singapore to start robotics. So that’s how I got involved in robotics. I’m still doing it now.
If you look at the most accessible application of robotics, it’s the autonomous vehicle. After all, everybody needs transportation. For example, it would be nice if even if you’re old and unable to drive anymore, you can still visit your friends anytime. You don’t have to call your kids to bring you around, you’re independent. AVs make that possible.
Even if you’re too young to drive, you would still have mobility. Whether you’re rich or poor, old or young—you need to go places, and self-driving cars allow you to do that.
They can ease congestion because AVs drive in the most efficient way. They won’t suddenly change lanes. Most of the congestion is caused by erratic human behavior, because [humans] are always in a hurry. If there are still human-driven cars, however, AV’s effect in easing congestion will be minimal. If you really want to ease congestion, all cars must be AVs.
But the main reason for AVs isn’t for easing congestion, it’s for safety. An AV has sensors that see everything, and they also know how far things are. Humans can’t do that. Drivers can see obstacles but they can’t calculate the exact distance, they only know near or far. Finally, computers don’t get tired or emotional. They don’t fall asleep.
The first challenge is predicting human behavior to drive more intelligently. The second challenge is localization. Right now we rely on GPS, but it isn’t very precise nor does it work indoors. Sensors also don’t work in extreme weather conditions like rain and snow. So there’s still remaining work to be done before AVs become mainstream
Singapore, however, is very aggressive. We are aiming to have AVs more widely available in three years. There will be three areas, namely Punggol, Tengah and Jurong Innovation District, with autonomous public transportation centered around train stations. Because in Singapore, AVs aren’t for long-distance highway driving, they’re for short, fixed routes. In fact, there’s already a driverless bus traveling from Jurong Island to Boon Lay. But the bus still has a human safety driver, who is prepared to take over if anything happens.
The first issue is whether robots will take away the jobs of humans. My answer to that is: if robots replace humans for these jobs, then that’s okay. If a machine can do what you’re doing, that means you should do something else that machines cannot do. The jobs that machines can do likely aren’t meaningful; they’re mundane and repetitive.
Another ethical issue is defining humanity. Say in the future, there’s a robot that is almost indistinguishable from humans. Does that robot have the rights of a human being? If that robot kills somebody, can you jail that robot? Should it stand trial? Or are you wasting taxpayers’ money? Should we just turn it off and destroy it? These are some important ethical questions that we need to ask.
I’m working on a big project on human-robot collaboration: specifically, creating the ultimate robotic assistant. A robotic assistant means not only building the intelligence, but also its physical body. This is why I’m very interested in soft robotics. After all, I don’t want huge, scary factory robots to be in my home. We need something soft and cuddly.
I also have many small projects with practical applications. For example, a robot that can climb trees for inspection. I’m also working on a mobile robot that can clean public toilets, as well as robots that can pick up utensils and clean tables at hawker centers.
I am excited about making robots intelligent in the same way that humans are. How can you capture the way humans learn and put it in a robot? With humans, you cannot accelerate learning. If it takes six years, it takes six years through experience. But with computers, that six years can be condensed into one hour. I’m interested to understand how humans learn and replicate that in machines.
First, give them an Internet connection so they can learn. There are many free online tools for building robots. If they’re in a rural area, give them a simple computer so that they can have internet access. There are also tools to easily build hardware, like Arduino. Also show them what can be done and excite them. So just spark their interest, give them a computer with internet access—they should then be able to do robotics.
This article is from a monthly series called Asia’s Scientific Trailblazers. Click here to read other articles in the series.
Copyright: Asian Scientist Magazine.
Disclaimer: This article does not necessarily reflect the views of AsianScientist or its staff.