The 10 challenges facing the robot future
Add Axios as your preferred source to
see more of our stories on Google.
At the National Institute of Advanced Industrial Science and Technology in Tokyo. Photo: Tomohiro Ohsumi / Getty
The current apex of useful commercial robots is a vacuum cleaner, and much work is left before machines can assume a more central role, from creating new basic materials to adopting ethical rules for their use, suggests a paper published today.
What's new: Guang-Zhong Yang, a professor at Imperial College London, led a global survey to ferret out the major remaining hurdles facing the field of robotics. The result is a list of 10 challenges (below).
The bottom line: Study co-author Manuela Veloso tells Axios that ultimately humans will be in control of how robots operate and the role they play. "These robots did not come from Mars and fall on Earth. They were invented by us and they will continue to be invented by us," she said. " ... [R]obots can have bad uses and good uses, and [this is] a call to people to make it right, to do it right."
How the survey was conducted: Yang and 16 co-authors conducted an online survey of researchers from various disciplines around the world. They vetted the results, short-listed 30 of them, and grouped them into the following list of 10.
- New materials will be needed to build robots that can complete new tasks. Scientists working with soft robots are at the forefront of this challenge, using squishy and bendable materials to develop robots that have self-healing skin and can contort to fit into cramped spaces.
- Bio-inspired robots derived from these new materials could have capabilities that mirror natural phenomena. Researchers at U Penn, for instance, have developed a material that transforms from a 2D sheet to a 3D shape, adjusting its texture to blend with its surroundings. The material mimics the camouflage capabilities of octopuses.
- Finding new power sources and batteries may allow robots to last longer and be mobile.
- Teamwork, or “robot swarms,” could add versatility. The ideal: a big robot that breaks into smaller operational units. Scientists in Brussels, for example, have designed modular robots that merge their control systems and become a larger entity.
- Robots that can adapt to their environment and its challenges, such as damage inflicted on the job, would be much more useful. Researchers created malleable elastomers that heal themselves after getting cut; and at Stanford, researchers created soft robots that self-inflate, growing or shrinking to fit their surroundings.
- The ability to learn on the job is another must-have toward making robots more human-like, says Veloso.
- Brain-computer interfaces are one way to connect robots to humans, so they can learn from us.
- Teaching social interaction. Robots won't be able to function alongside humans until they can "understand human social dynamics and moral norms," the study says.
- Using robots in medicine. The future of medical robots is autonomous entities that heal us. Another frontier is swallowable micro-machines that enter our bodies and treat us. Harvard roboticists made a breakthrough in this area by developing tiny bots that can fold and unfold themselves.
- Ethical questions. Once a robot becomes capable of making decisions, the ethics of its judgement inevitably come into question, says Veloso. 'We would like robots to use principles and have values," and this involves collaboration with philosophers and psychologists, she says.
