Skip to main content

Dario Floreano and Nicola Nosengo - Five Way Interview

Dario Floreano is Director of the Laboratory of Intelligent Systems at the Swiss Federal Institute of Technology Lausanne (EPFL). He is the coauthor of Evolutionary Robotics and Bio-Inspired Artificial Intelligence (both published by the MIT Press). Nicola Nosengo is a science writer and science communicator at EPFL. His work has appeared in Nature, the Economist, Wired, and other publications, and he is the Chief Editor of Nature Italy. Their recent book is Tales from a Robotic World.

Why robotics?

Robotics is where Artificial Intelligence becomes tangible and personal. You may forget about the AI that powers your online searches, but it is hard to ignore the physical presence and motion of robots next to you. Robots have transformed factories, they are widely used in warehouses, hospitals, even homes - at least in the form of vacuum cleaners. Drones are becoming ubiquitous. And yet there is so much more that robots could do for us, and so many more places where they could prove useful, if they were more autonomous and capable of learning from their experiences and from us, if they could understanding the meaning of what they see and what they hear, if their mechanical bodies could be more adaptable and life-like. Thanks to the convergence of engineering, neuroscience, materials science, robotics is now entering a new phase, and that makes it one of the most exciting fields of science now.

Why this book?

Surprisingly, physical robots are largely neglected in the media and in books. Often, when people talk or write about "robots" nowadays they really talk of software-based AI, algorithms and data. Indeed, AI has had amazing successes lately, and it tends to overshadow physical robotics, which may seem to lag behind. In fact, making robots intelligent is harder than making intelligent data-crunching algorithms, but there are many recent advances that will make robots more capable and pervasive in the near future, and that deserve to be known by the public. With this book, we wanted to fill this gap and to show where robotics research is now. But we also understand that people want to know how all of that research will make a difference in their lives, which is why we stuck our neck out and described concrete future scenarios made possible by current-day research.

Despite all the promises, robots (and even driverless cars, particularly on winding European roads) always seem to be behind promised timescales. Why do we underestimate how long it takes for these technologies to mature?

Part of the reason is that, when it comes to predicting the pace of technological progress, we have been spoiled by what happened in computation, where computing power has grown exponentially according to what is usually called Moore's law, and the performances of all machines that rely on microchip have grown as a consequence. Many people have made the mistake of thinking that progress in robotics is a matter of increasing computing power, as it is the case with computing. But technologies that work with the physical world are more complicated than those that work with data - and that includes driverless cars, that are in effect robots that have to move in unpredictable environments and among people. They require not just improvements in algorithms and computing chips, but also in material science, mechanical engineering, sensor development, social interaction, cognitive science, and much more. Creating a new generation of robots requires new design and control principles, and a much closer collaboration with biologists, neuroscientists, and psychologists than in classical engineering or computer science. 

What’s next?

It would be fun and instructive to turn the book into a documentary series or a movie. Robots deserve to be seen live!

What’s exciting you at the moment?

N. I don't know if 'exciting' is the right word, but the conversion of our economy away from fossil fuels and towards renewable energy is definitely the defining challenge of our era, one that will require a lot of new research combined with a lot of social innovation. It will be exciting to watch it happen over the next couple of decades, hoping that there is enough political will for it. By the way, robotics will have to be part of this transformation. We can't flood the world with yet more machines and make them energy-hungry, or harmful for the environment. The robots of the future will have to be sustainable, or they won't be at all.   

D. My research projects include edible robots and robotic food; bird-inspired robots capable of moving in the air, ground, forests, and water; human augmentation by robotic swarms, and evolutionary robots.

Interview by Brian Clegg - See all of Brian's online articles or subscribe to a digest free here

Comments

Popular posts from this blog

The Language of Mathematics - Raúl Rojas ***

One of the biggest developments in the history of maths was moving from describing relationships and functions with words to using symbols. This interesting little book traces the origins of a whole range of symbols from those familiar to all, to the more obscure squiggles used in logic and elsewhere. On the whole Raúl Rojas does a good job of filling in some historical detail, if in what is generally a fairly dry fashion. We get to trace what was often a bumpy path as different symbols were employed (particularly, for example, for division and multiplication, where several still remain in use), but usually, gradually, standards were adopted. This feels better as a reference, to dip into if you want to find out about a specific symbol, rather than an interesting end to end read. Rojas tells us the sections are designed to be read in any order, which means that there is some overlap of text - it feels more like a collection of short essays or blog posts that he couldn't be bothered ...

Why Nobody Understands Quantum Physics - Frank Verstraete and Céline Broeckaert **

It's with a heavy heart that I have to say that I could not get on with this book. The structure is all over the place, while the content veers from childish remarks to unexplained jargon. Frank Versraete is a highly regarded physicist and knows what he’s talking about - but unfortunately, physics professors are not always the best people to explain physics to a general audience and, possibly contributed to by this being a translation, I thought this book simply doesn’t work. A small issue is that there are few historical inaccuracies, but that’s often the case when scientists write history of science, and that’s not the main part of the book so I would have overlooked it. As an example, we are told that Newton's apple story originated with Voltaire. Yet Newton himself mentioned the apple story to William Stukeley in 1726. He may have made it up - but he certainly originated it, not Voltaire. We are also told that â€˜Galileo discovered the counterintuitive law behind a swinging o...

Ctrl+Alt+Chaos - Joe Tidy ****

Anyone like me with a background in programming is likely to be fascinated (if horrified) by books that present stories of hacking and other destructive work mostly by young males, some of whom have remarkable abilities with code, but use it for unpleasant purposes. I remember reading Clifford Stoll's 1990 book The Cuckoo's Egg about the first ever network worm (the 1988 ARPANet worm, which accidentally did more damage than was intended) - the book is so engraved in my mind I could still remember who the author was decades later. This is very much in the same vein,  but brings the story into the true internet age. Joe Tidy gives us real insights into the often-teen hacking gangs, many with members from the US and UK, who have caused online chaos and real harm. These attacks seem to have mostly started as pranks, but have moved into financial extortion and attempts to destroy others' lives through doxing, swatting (sending false messages to the police resulting in a SWAT te...