In addition to its use in investigating the possibility of extra-terrestrial life, aerospace robotics technology is also being applied to the fields of self-driving cars and pilotless planes. We speak to Marco Pavone, a Professor at Stanford University recently honoured by President Obama for his research into space robotics, which are set to “be increasingly, autonomous, safe and useful.”
It may appear difficult to imagine the link between space robotics and driverless cars but for those who know about such things, such as Marco Pavone, the connection is real and the subject of valuable work.
The Italian engineer, born in 1980 and a graduate of Catania University, worked at the Jet Propulsion Laboratory (JPL), a research facility for NASA, before arriving at Stanford University in California, where he is assistant professor of aeronautics and astronautics. He recently received a Presidential Early Career Award for Scientists and Engineers from Barack Obama for his research and the results obtained in the field of aerospace systems. It is on this topic that our conversation begins before moving on to the subject of driverless cars. At Stanford Pavone is involved in research into both.
Aerospace systems: what are they and what are they for?
There are missions into the Earth’s orbit and missions beyond the Earth’s orbit. The former have both a scientific and technological value. The scientific value because it is possible to study the Earth in a unique way, for example, the atmosphere and the oceans. For the technological questions, think of GPS, and its European variant (Galileo) or the Russian version (GLONASS). Then there’s the exploration of the solar system. In this case the value is prevalently scientific, from the study of the origin of the sun and other planets to the investigation into the possible existence of alien life forms.
Up to this point everything is clear. Now let’s move on to one of the most important innovations in the field of aeronautics in recent years: the miniaturization of satellites and space systems in general and the consequent fall in the production costs to make them. Today, with relatively modest sums it is possible to launch a small satellite into orbit. In Silicon Valley there are a number of different start-ups working in this very field. Of course, that does not mean that the big satellites that are difficult to build and more costly are no longer necessary but soon they will probably be accompanied by fleets of mini satellites that could, for example, monitor the Earth continuously.
Space robots represent a further step. These could be used to resupply costly satellites with fuel or to assemble complicated structures directly in space. Furthermore, the space robots are fundamental for exploring other planets, from Mars to asteroids and comets. In particular we are developing, in collaboration with JPL, a robot to explore micro gravitational environments (for example, asteroids) capable of exploiting the movement of an internal mass in order to “jump” from one point to another on the surface (video).
Why the interest in exploring asteroids and comets?
By studying them it is possible to understand how the solar system evolved and it’s even possible to try to comprehend the likelihood of life beyond Earth.
And is there any?
There is hope or fear according to your point of view. In the solar system elements considered precursors to life have already been found. Furthermore, recently diverse “exoplanets” (planets that do not belong to our solar system,) have been discovered: these are optimal candidates for hosting life forms. The study of these could further corroborate the idea that we are not alone in the universe.
So, how do we get from space robots to driverless cars?
From a technical point of view, in my laboratory we work on artificial intelligence algorithms for the control of autonomous robotic systems. With a bit of adaptation these algorithms can be used either for space robots or for driverless cars. Obviously there are differences but in both cases we’re talking about autonomous systems that must be capable of making decisions to deal with unexpected situations.
So, are driverless cars almost with us?
That depends what you intend by driverless cars. I believe that autonomous shuttles that operate in “controlled” scenarios such as parks or limited traffic zones are not far away. On the contrary, for autonomous cars that can operate in any traffic conditions it will take another ten years if not more.
Will autonomous robots take jobs away from people?
In difference to the automation that took place in factories and agriculture gradually substituting jobs that were prevalently repetitive and enabling work to regenerate and find other outlets, autonomous robots could rapidly replace various types of cognitive work, thus potentially creating some problems. For example, driverless cars will be able to drive on motorways with relative ease, therefore, jobs linked to transport and logistics could be hurt by these developments in a rapid and substantial way.
Bill Gates is proposing a tax on robots. Is there any sense in that?
Prior to the industrial revolution people would work twelve hours a day and sometimes more. Technical developments enabled a reduction in working hours. Might working hours reduce even further to perhaps 20 hours per week? If the answer to that is yes then we must ask ourselves whether this is socially acceptable as well as economically possible. Taxing robots is an option but it is also possible to imagine new social welfare systems. The concept of a universal basic income is being discussed even here in the USA, the country of capitalism. There are many studies on its potential impact. In Stanford I’m working with sociologists, economists and political scientists on this very subject. The university is one of the best places to carry out this type of research because we don’t have conflicts of interest (for example of a financial nature). I would add that for those of us who design robots it is our duty to think about the social consequences that our inventions could provoke.