Living with Robots: Fiction, Reality or Opportunities?

Robotics technology is currently in a golden age, with robots improving our quality of life and helping with pressing social problems. In this edition of the Interdisciplinary Distinguished Speaker series, held on 25 November 2021, Professor David Hsu, Provost’s Chair Professor, Department of Computer Science, NUS talked to us about recent developments in, unrealistic expectations about and the future of robotics.

Current developments

Professor Hsu began the session sharing some current state-of-the-art robotics examples, including GoAnywhere@NUS, a robot that can move and adapt to its changing surroundings. Beyond NUS, there is also BigDog, which can navigate difficult terrains, AI skin that allows robots to pick up something as fragile as a piece of tofu, and the da Vinci surgical system that aids in remote controlled surgeries. In more commercial settings, Amazon’s ecommerce shelves have base robot platforms that help improve work efficiency.

Addressing these different types of technologies, Professor Sharon Ng, Head of Marketing Division, Nanyang Business School, NTU, wondered if the robotics industry would be fragmented or dominated by big players. Professor Hsu suggested that it would be both as robotics is a very sophisticated technological field but the industry is also very broad. Currently, breakthroughs happen on both ends of the spectrum where, on one end, start-ups work on various components of a robot (e.g., eyes, hands, brain, etc.) while big players like Google work on base platform capable robots.

Professor Xin Chang Simba, Associate Dean (Research), Nanyang Business School, NTU followed with a question on financing the robotics industry. Technological transformation has led to huge technological development, Professor Hsu said, but real-world, large-scale deployment is still not present. One reason is that a long-term perspective for robotics technology is important—to produce something that is as capable and reliable as BigDog, for instance, would require at least 25 years of hardcore research and another 20 years of technological development. Having said that, Professor Hsu suggested that investment committees do understand this. For instance, BigDog creator, Boston Dynamics, has not really made money but has been kept alive by investors for the past 20 over years.

Misconceptions about robots

Much of our understanding about robots stem from the movies but Professor Hsu stressed that most movies are wrong. There are many examples of advanced robots that fail our expectations because (1) the technology is deployed in an inappropriate manner, and (2) robot demonstrations, particularly in a video, often show a robot in a specific performing condition. However, the success of a robot, Professor Hsu emphasised, depends on the variability of the environment and how well the robot adapts to it. Consequently, we are bound to be disappointed if we were to base our decisions only on such demonstrations.

In particular, Professor Hsu highlighted the tendency to underestimate the importance of variability and adaptivity, and to overestimate the importance of speed, strength and precision in robots. Referring to a poll conducted at the beginning of the webinar that showed around 80% of participants would prefer a human receptionist to a robot one, Professor Hsu suggested that this is probably because the robot receptionist is currently not good enough. What is needed to pass consumers’ threshold is not available yet as tasks that are simple to humans can be very complex for a robot. Furthermore, acceptance of robots also depends on its relative price: In a hospital, for instance, where every medical equipment is expensive, investing in a robot may be more acceptable.

In order to change people’s perception about robots, Professor Hsu suggested that it is crucial for designers to ensure that the values given to the robots align with our own values. Once that is achieved, the chances of engendering trust and encouraging adoption would be higher.

Serving social needs

While robots were largely used in manufacturing settings in the past, the new generation of robots could help with issues such as an aging population. Robots can help reduce human labour, increase productivity, and create safer working environments. Rather than replacing humans workers, robots help to expand human capabilities.

Consequently, the new generation of robots need to interact and collaborate with humans seamlessly. Robots of the future need to work in close proximity with humans and, thus, must adapt to a wider variety of situations. Designers will need to consider issues such as soft materials to ensure safety and communication technology for humans and robots to interact with each other.

China and robotics

Professor Hsu also drew specific attention to China, pointing out that China’s deployment of industrial robots has grown at an exponential rate from 2016, making it the highest net consumer of robots as well as a powerhouse of new robotics technology. In particular, China is moving very fast in absorbing advanced robotics technology and turning them into products and services, while inventing new technologies themselves at the same time.

While there may be concerns that the deployment of industry robots may lead to social issues such as unemployment, Professor Hsu thinks that this may not happen in China since the increased deployment stems, in part, from the needs of an aging workforce.

The Singapore context

On STEM education in Singapore, Professor Hsu feels that experts need to interact with each other to widen perspectives. Robots need both computer and mechanical components, and individuals that have expertise in one area need to remain aware of the other. Professor Hsu suggested that Singapore currently lacks talents who can connect the various components of a robot.

With regards to the use of drone for delivery in Singapore, Professor Hsu pointed out that this technology might be available within five years. However, he also suggested that this requires both a technological and business solution. While delivery to the door may be a huge technological challenge for drones, it is possible to deliver packages to collection points. This, however, would require organisation and a consideration of business and legal issues for this to become a reality.