Elon Musk’s recent announcement on Twitter that “Tesla will have some really useful humanoid robots in production in small numbers for internal use next year” suggests that robots with human-like physical characteristics and “really useful” capabilities could be coming soon.
Yet despite decades of attempts, useful humanoid robots have remained a fiction that never seems to match reality. Are we finally on the cusp of a breakthrough? It’s worth asking whether we really need humanoid robots.
Tesla's Optimus robot is just one of several emerging humanoid robots, joining the likes of Boston Dyanmic's Atlas, Figure AI's Figure 01, Sanctuary AI's Phoenix, and many others.
They usually take the form of a bipedal platform that is capable of walking to varying degrees, and sometimes jumping, along with other athletic feats. Atop this platform may be a pair of robotic arms and hands that are capable of manipulating objects with varying degrees of dexterity and touch.
Behind the eyes is artificial intelligence adapted to plan navigation, recognize objects and perform tasks with those objects. The most common applications for such robots are factories, performing repetitive, sullied, uninteresting and unsafe tasks, and working with humans in a cooperative manner, such as carrying a ladder.
They are also proposed for work in the service sector, possibly replacing the current generation of more practical service robots such as “greeters” or “tour guides”.
They can be used in social care, where attempts have been made to lift and carry people, like Riken Robear (unfortunately resembling a bear rather than a humanoid), as well as in providing personal care and therapy.
There is also a more established and developing market for humanoid sex robots. Interestingly, while many see moral and ethical issues associated with them, the utilize of humanoid robots in other areas seems to be less controversial. However, it turns out that delivering humanoid robots in practice is arduous. Why is this?
There are many engineering challenges, such as achieving malleable bipedal locomotion on different terrain. It took humans about four million years to do that, so where we are now with humanoid robots is pretty impressive. But humans are learning to combine a convoluted set of sensing capabilities to achieve that feat.
Similarly, achieving dexterous manipulation of objects that come in all shapes, sizes, weights, and levels of fragility has proven stubbornly arduous for robots. But there have been significant advances, such as the dexterous hands of British company Shadow Robot.
Compared to the human body covered with pliable and malleable skin that constantly perceives external stimuli and adapts to its environment, robots' tactile abilities are narrow to only a few points of contact, such as the fingertips.
Moving beyond automating specific tasks on factory assembly lines to improvising general tasks in a lively world requires greater advances in AI, as well as sensing and mechanical capabilities. Finally, if you’re going to make a robot look like a human, there’s an expectation that it will have to communicate with us like a human, and maybe even respond emotionally.
But this is where things get complicated, because if our brains, which have evolved to recognize nonverbal elements of communication, don’t perceive all the micro-expressions interpreted on a subconscious level, a humanoid robot can seem downright terrifying.
These are just a few of the stern research challenges that researchers working in the field of robotics and human-robot interaction around the world are already facing.
There is also the additional limitation of deploying humanoid robots in our ever-changing, clamorous real world, with rain, dust, and heat. These are completely different conditions than those in which they are tested. So shouldn’t we focus on building systems that are more resilient and don’t fall into the same pitfalls as humans?
We recreate ourselves
This brings us to the question of why Musk and many others are focused on humanoid robots. Do our robotic companions have to look like us? One argument is that we have gradually adapted our world to the human body.
For example, our buildings and cities are largely built to accommodate our physical form. So it is an obvious choice for robots to take on that form as well.
It must be said, however, that our built environments and tools often assume a certain level of strength, dexterity, and sensory abilities that puts a huge number of people, including those with disabilities, at a disadvantage. So would the proliferation of more powerful metal machines among us further perpetuate this divide?
Perhaps we should see robots as part of a world we need to create that better accommodates the diversity of human bodies. We can put more effort into integrating robotic technology into our buildings, furniture, tools, and vehicles, making them smarter and more adaptable, so they are more accessible to everyone.
It is striking how the current generation of narrow robot forms does not reflect the diversity of the human body. Perhaps our apparent obsession with humanoid robots has other, deeper roots.
God's desire to create a version of ourselves is a fantasy that appears repeatedly in dystopian science fiction novels, ideas from which the technology industry eagerly borrows.
Or perhaps humanoid robots are a “moonshot,” a vision we can all understand but that is incredibly arduous to achieve. In brief, we may not be entirely sure why we want to go there, but impressive engineering innovations are likely to emerge from the attempts themselves.
Steve Benford, Professor of Collaborative Computing, University of Nottingham and Praminda Caleb-Solly, Professor of Embodied Intelligence, Department of Computer Science, University of Nottingham
This article was reprinted from Conversation under the Creative Commons license. Read original article.
Image Source: Pixabay.com
