Robotic engineer Subramanian Ramamoorthy will outline his vision for the future during his talk at New Scientist Live at ExCel London on Saturday, Intelligent robots: friend or foe? He told Express.co.uk: “In recent years, technologies underpinning robotics and AI have developed rapidly and come together with corresponding improvements in hardware – computing and mechanical – to enable entirely new applications. “So, we have seen autonomous cars go from being laboratory curiosities to business units in car companies, we see robots performing numerous tasks in warehouses and hospitals, etc.
“A common theme emerging in all of these applications is the need for reliability; now that people are seriously such autonomous systems, we are keen to understand how and when they fail.
“This has been trickier than one would have imagined, in part due to the ways in which we have come to this level of technical ability – through the use of data-driven learning, online adaptation, and combination of complex pieces that have to parallel the infinite richness of the world in which the robot operates.”
Virtual assistants such as Apple’s Siri and Google’s Alexa are becoming ubiquitous – but Mr Ramamoorthy said he did not classify them as robots.
He explained: “Programs like Siri don’t really do much by way of action (unless they can persuade the user to do something – in which case they would be doing something interesting for robotics), and they are not particularly interesting in terms of making decisions (they mostly follow very limited scripts).
Robots could be assisting in operating theatres within a decade, said Mr Ramamoorthy
Boston Dynamics’ robot “dog” Spot meets a real canine
“For me, better examples include autonomous warehouse and delivery machines that take high level guidance and execute movements in an intelligent way, or robots that can pack your groceries into a shopping bag.
“Personally, a direction I am very excited by is the use of robots in the operating theatre, to help surgeons with tasks where sensor-guided motion can help them achieve improved outcomes.
“I am hopeful that we will begin to see such developments within a decade, or two.”
Mr Ramamoorthy said in order to achieve a “useful level of safety”, it would be necessary to weave together many threads of research activity.
Subramanian Ramamoorthy will be at New Scientist Live on Saturday
Nor is there any guarantee that robots would end resembling human beings, with two arms, a leg and a head.
Mr Ramamoorthy said: “It is possible. There are already some such robots out there. For instance, Boston Dynamics have just released a commercial version of “Spot” – it is not human like, but it is certainly very nimble like our four-legged animal friends.
“The engineer in me says that mimicking the human body for its own sake is likely unnecessary. Instead, we should think about what that body enables, and how we can borrow the essential principles.
“For instance, we may decide that we want the agility of a mountain goat – and build something like Spot to mimic that. Or we may decide we want the lightness of touch that our hand exhibits, and we could copy that without the baggage of the rest of our body.”
Jet suit breakthrough: ‘Buck’ Rogers in the 21st Century [VIDEO]
Oumuamua news: Do intergalactic nomads really ‘seed the universe’? [ANALYSIS]
What lies beneath? Oceans ‘could hide undiscovered giant octopuses’ [PICTURES]
Prince William interacts with a robotic limb
Subramanian Ramamoorthy said robotics will play an important role in future space exploration
As robotics technologies become increasingly sophisticated, and the debate about artificial intelligence, self-awareness and even sentience become increasingly fierce, Mr Ramamoorthy admitted philosophical and ethical dilemmas lie ahead.
He said: “This is a difficult question mainly because there is debate, within philosophy and in the abstract, about what ‘self-awareness’ means.
“It is certainly the case that machines will become incrementally better at learning, and through that achieving increasingly better models of the world and of themselves.
“Whether this will one day take us into the realm of such models being akin to consciousness (whatever that may be, even in humans) is a topic best left for speculation, based on current evidence.”
New Scientist Live is at ExCel London from October 10-13
As for concerns about robots taking jobs previously done by human beings, he added: “The question of replacement is more interesting. They could do so in a trivial sense, in the same way that we no longer have typist pools, and in the way most of us use washing machines to take care of our laundry.
“But it will be a long time before one integrated machine possesses sufficient capability in built within it, to be a meaningful competitor to the full person.”
As for mankind’s aspirations for space exploration, Mr Ramamoorthy believes robotics will be key.
He said: “The main advantage of robots, when it comes to interplanetary exploration and travel, is that they do not suffer so much from the time barrier of vast interstellar spaces.
A robotic, prosthetic arm in action
“For instance, it will take years to get to our nearest planetary neighbour and would likely involve going through treacherous magnetic fields and storms.
“Robots could be built to handle this while it is risky to expose humans to this (and it is not even clear if humans could bear the mental stresses, while robots don’t face that problem).
“So, robots really are a tele-present extension of ourselves and that is the role in which they will be very useful for the near future.
“As they become more capable, and agile in body and thought, we can envision a world where we task them with high level goals and send them away, with instructions to report back if they find useful things!”
Intelligent robots: friend or foe? is on the Technology stage on Saturday between 1.45pm and 2.25pm