The field of robotics is accelerating at an increasingly rapid rate and is showing no signs of slowing down. But there remain some significant robotics challenges standing between us and real progress.
Recently, a team from the journal Science Robotics conducted an open online survey to better understand some of the major unsolved robotics challenges – aka what keeps roboticists up at night?
Following the survey, they created a panel of 10 experts who shortlisted the 30 most important topics and research directions. These were then grouped into 10 major robotics challenges that “might have major breakthroughs, significant research, and/or socioeconomic impact in the next five to 10 years”.
Many of these challenges build on each other and require collaboration and innovation to solve them. But the benefits are manifold, as these challenges – and resulting solutions – will transform almost every facet of our lives in the decade, from medical care, to exploration to offices.
1. New materials and fabrication schemes
While robots have made significant advancements in recent years, they are still made from bolts, bearings and motors. These rigid materials make them unstable on uneven surfaces and often means their contact with humans is limited to avoid injury. For example, factories that employ both human and robot workers often separate them by a physical barrier to prevent potentially dangerous collisions.
Researcher and report contributor Steve Davis said this is problematic because it prevents “even the most basic of cooperation”. He suggested robots made out of more lifelike materials could transform the experience of human factory workers.
“A robot might do the hard work of supporting a heavy component while the human performs a complex assembly task on it. The machine provides the brawn and the person provides the brains, significantly improving what human-robot collaboration currently can achieve,” he said.
Others who contributed to the report point to the contribution ‘soft robots’ could make to the medical field.
“I could see soft robots being especially useful for tasks where you use fingers, like surgery,” said contributor George Whitesides of Harvard University.
However, the development of these robots is far from straightforward. Science Robotics said developers must still overcome basic hurdles, including “improved portable energy storage and harvesting, new materials with tuneable properties, and new fabrication strategies to embody these functional materials as new capabilities for future robots, including the robot building and repairing itself.”
2. Biohybrid and bioinspired robots
Connected to the idea of soft robots are robots inspired by the animal kingdom, also known as biohybrid and bioinspired robots.
“By combining robotics with tissue engineering, we’re starting to build robots powered by living muscle tissue or cells. These devices can be stimulated electrically or with light to make the cells contract to bend their skeletons, causing the robot to swim or crawl,” said mechanical and aerospace engineering Victoria Webster-Wood in an article for The Conversation.
However, the Science Robotics report found that major challenges remain for nearly all component technologies that could enable bioinspired behaviour.
“Materials that couple sensing, actuation, computation, and communication are critical and must be shared as developed,” the report stated.
3. Power and energy
“As for any electronic system,” researchers said, “power and energy sources represent one of the most challenging areas of robotics research.”
This is particularly important for those hoping that robots might be able to work in industries that require them to be in extreme environments, such as deep sea exploration.
“Beyond the currently available commercial technologies such as lead-acid, nickel–metal hydride, and lithium-ion batteries, there has already been extensive research on developing next-generation technologies, such as fuel cells and supercapacitors,” Science Robotics reported.
Advancements in biohybrid robots could advance this area.
“No battery can yet match the metabolic energy generation in organisms,” the report stated, but “biohybrid robots could use the unique features of living cells for potential solutions.”
4. Robot swarms
The concept of robot swarms is also derived from nature. In the same way that birds, fish and insects move in cohesive groups, researchers are working towards developing robots that can act as part of a team to achieve complex goals.
If successful, Science Robotics predicts these swarms have the potential to solve the most pressing problems facing human civilisation.
“They can provide solutions to feed an ever-increasing population with limited resources by increasing the efficiency of food production and decreasing water consumption by an order of magnitude. They can respond to natural disasters and adversarial attacks by enabling resilience in our infrastructure. They are a part of any practical solution to space colonisation,” the report stated.
But the development of these swarms is complex. Individual robots need to sense not only the environment but also the movements of their neighbours, all while communicating with other individuals in their teams and acting independently.
At present, robots are able to function autonomously using a perception-action loop. This feedback loop is fundamental to creating autonomous robots that function in unstructured environments.
Robot swarms, Science Robotics said, “require their communication ability to be embedded in this feedback loop. Thus, perception-action-communication loops are key to designing multifunctional, adaptive robot swarms. There are currently no systematic approaches for designing such multidimensional feedback loops across large groups.”
5. Navigation and exploration
One area in which robotics is poised to make a major contribution is exploration, particularly in extreme environments like deep sea, terrestrial and arctic conditions that are dangerous for humans. Researchers also hope that robots might be able to navigate and undertake nuclear decommissioning tasks.
According to Science Robotics this means current navigation systems will continue to evolve, allowing robots to be increasingly autonomous. They will also require “the physical robustness to withstand harsh, changeable environments, rough handling and complex manipulation.”
Self-reconfiguration and repair will also be important for robots in extreme conditions.
“When possible, solutions need to involve control of multiple heterogeneous robots; adaptively coordinate, interface, and use multiple assets; and share information from multiple data sources of variable reliability and accuracy,” Science Robotics reported.
6. AI for robotics
The development of complex intelligence in robots is somewhat inhibited by our lack of knowledge about the human brain – specifically, our ability to map the neocortex.
According to Science Robotics, a recent paper has provided “some detailed and testable predictions concerning how columns in the neocortex provide location signals that enable learning the structure of the world.”
However, more testing needs to be done to fully understand this structure and to recreate it using technology.
One of the great tests for AI remains ethics, and researchers are continuing to work to create machines with complex moral reasoning skills. Science Robotics predicts that we will see “considerable and rapid progress on this front”.
7. Brain-computer interfaces
Brain-computer interfaces, also known as BCIs, have already been transformative in healthcare, translating a brain activity into physical action in patients suffering from illnesses impacting motor function.
As this technology is usually worn by its users, a priority in this area is the development of materials that will improve patient experience. To that end, microfabrication, packaging and flexible electronics, along with local processing and wireless data paths are all areas that researchers will continue to explore.
Researchers are also looking to make the technology more agile, accounting for variation among different brains. The way in which this challenge is currently met is not sustainable, according the reports authors.
“Current methods often involve extended periods of training, calibration, learning and adaptation, thus making it prohibitive for general use,” they said.
8. Social interaction
One of the more complex challenges for robotics is social interactions. Researchers said this skill is vital for integrating robots into human environments, including schools, hospitals, shops and homes.
“Humans are so adept at recognising and interpreting social behaviour, we often underestimate the complexity of the challenge that this represents for a robot” they said.
Furthermore, because we are so good at social interaction, few comprehensive, quantitative studies have been done in this area that might better enable engineers to program robots to understand the subtleties of human interaction, particularly across cultures and demographics.
“Next-generation systems will need to richly mix elements from multiple input modalities and combine these perceptual systems with predictive models of social intention to more fully capture the rich, dynamic nature of social interactions,” they said.
9. Medical robotics
However, their domination in this sector is not without its challenges. Researchers hope that AI enabled robots will be able to work alongside doctors and surgeons, but also that robotic technology could be used as dynamic implants in the place of donated organs.
For both these tasks, materials are of the utmost importance. As well as the development of soft robots, which will create a safer working environment for both medical staff and robots, challenges also include biocompatibility, reliability, adaptability, security and providing power.
10. Robot ethics and security
As robots become increasingly integrated into society, researchers noted the importance of ethics and security.
They said that both institutions and engineers must account for the complicated economics of introducing robots to the workforce, as well as ensuring that the machines behave ethically.
Contributors also cited the need to maintain skilled professions in roles that might become dominated by AI, but will nevertheless still require some degree of human oversight.
“Radiologists need to keep studying images for the same reason pilots need to keep landing airplanes: so that they still can even if the AI cannot, or if the AI gets it wrong,” they said.
Engineers have also been confronted with the problem that technology can be used malevolently if it finds its way into the wrong hands.
“Examples range from scanning citizens’ faces in illiberal regimes to discriminating among applicants for a job or punishing law offenders unfairly,” they said.
Work in this area has already begun, although there has been little consensus. Contributors refer to the ethical philosopher, Kant, suggesting that “AI should be designed and used to treat every human being as always an end, and never only as a means”.
But it will ultimately be the shared responsibility of organisations, lawmakers and engineers to find ways to develop responsible parameters around the use of AI and robotics.