Floreano NCCR

On 15 April 2010, the EPFL Community got the news of the creation of eight new National Centers of Competence in Research (NCCR). The announcement, which emanated from the very official Swiss Confederations’ media centre, brought especially bright news to the EPFL robotics community (the NCCR Robotics Principal Investigators consists of 7 Professors and a Senior Researcher). Among them is Professor Dario Floreano, the Director of the Laboratory of Intelligent Systems (LIS), who federated the community around a coherent scientific and educational program in robotics. EPFL and its School of Engineering (STI) will be host to the new National Center of Competence in Research in Robotics under the directorship of Prof. Floreano.

Prof. Dario Floreano and his team learned about the news after having worked on the project proposal over a two-year period, ever since the initial call for project submission made by the Swiss National Science Foundation (SNSF). Under its mandate by the Swiss Federal government, the SNSF’s mission is to support research in all scientific disciplines, from philosophy and biology to nanosciences and medicine.

So what have Prof Floreano and his colleagues learned that day? They learned that they would be the pioneers in launching a Competence Centre dedicated to Robotics in Switzerland that, according to the SNSF, is “deserving of recognition for outstanding scientific merit.” Starting at the end of this year, the new centre under the directorship of Prof Floreano will receive funding of 13.3 million Swiss francs over four years. The University of Zurich, the Swiss Federal Institute of Technology Zurich (ETHZ), and the Istituto Dalle Molle di Studi sull'Intelligenza Artificiale (IDSIA) who also take part in the project, will act as partner universities. The function of a Competence Centre is to promote long-term research projects in areas of vital strategic importance for furthering of science in Switzerland, enhancing the Swiss economy, and deriving benefits for Swiss society.

Robotics: make something Swiss

When asked about what such recognition means for a scientist like him and for his career, Prof Floreano spoke of the common good for EPFL and the Swiss community of researchers involved in robotics and Artificial Intelligence (AI). This community shares the vision that robotics is a field of technology with both societal and economic challenges for the future. The new structure will help federate those who, until today, were clustered in different labs, in different institutes, or even in different areas such as micro-engineering, electrical engineering, or mechanical engineering.

Professors active in all areas of robotics and AI will now be able to team up and speak in a unified voice, especially when interacting with the international community, such as the similar technology centers in Japan and the United States. Students and the educational programs that participate in will also benefit from such outreach. “I thought it was a good time to bring all of us together”, commented Prof Floreano.

As part of the plan for the Competence Center in Robotics, a new building will house the Robotics section in EPFL, as well as two other areas of STI: Neuroprosthetics and Biomechanics. The new locale will be open to visitors who will be able to see students and researcher work on novel solutions using robotics. Prof Floreano and his team envision that this centre will provide a basis from which they can launch outreach efforts to other institutions, identify synergies, and establish collaborations with new partners.
Robots: helping humans rather than replacing them

Just as robotics is celebrating its 50th anniversary this year (festivities took place in July in Anchorage, Alaska, with almost 2000 researchers) it is settling down as a new discipline at EPFL. The new centre’s research program will focus on three core themes, with one motto: develop “human oriented robots”, a new generation of robots capable of facilitating our lives in many ways, and differently than what they were doing fifty years ago. At that time, robotics was focused upon assembling, producing, and packaging goods at a better quality and at a lower price. Robots were liberating workers from heavy, repetitive tasks in factories, at a time when the world economy was growing in the post-war II period faster than the available workforce. Although the need for industrial robots has not disappeared, robotics is taking on a new focus in helping humans face the new challenges of tomorrow.

“The first objective comes with aging societies,” explains Prof Floreano. “We want to remain active, and stay in our homes rather than live in senior living centers. Robots can help there and could be companions in daily life as well.”

But wait - even if people appreciate and live with robots at home in Japan, it is still quite different in western societies. “In Europe, people tend to prefer human company, rather than machines”, Prof Floreano explained. As research focus and resources will be allocated to robots in daily life, part of the NCCR Robotics' mission will be to investigate how robots are going to take a role in our society. The NCCR team will make interviews, deploy robots in houses and carry systematic experiments on how people interact with them, taking society’s feedback into account before developing new robots. “I hope that the sort of attitude and passion for robots that we see in Japan will come to western societies”, concludes Prof Floreano.

A second objective comes with urbanization and the increase of population density in cities. People tend to live in larger urban centers, creating both challenges and opportunities for robotics. For example: how to keep our urban environment well structured, monitored, and operating? The challenges such as: waste disposal, infrastructure repair and maintenance, air and water quality monitoring and control or the efficient transport of people and goods, can be addressed by robots that are ideally suited for such tasks. Finally, robotics will play a role not only in monitoring and preventing disasters, but also for search and rescue of people if a disaster occurs, and thus fulfilling the third objective in improving safety of life and emergency assistance.
Robotics: the Swiss competences

To deploy the technologies derived from the new centre's program, Prof Floreano can count on the competences of EPFL, ETHZ, the University of Zurich, and IDSIA, which will converge towards the fundamental breakthroughs needed in various areas in robotics. For example, on robots mobility: “until today, robots were mostly equipped with wheels, and adapted to a specific environment. In the NCCR we will develop jumping, flying, swimming, or snake-like motion, allowing our robots to move in cluttered environments, in the forest, among buildings, or in water".

Such competences will also benefit rehabilitation and prosthetics robotics, in collaboration with prosthetics scientists at EPFL, ETHZ, and the University of Zurich. For example, the NCCR team will work on the development of machines helping people to remain mobile for a longer period of time, for example wearable exoskeletons, allowing somebody at 75 to climb up a mountain! This is one of Prof Floreano’s vision for the future of prosthetics or rehabilitation robotics, as part of the centre.
Interview with Dario Floreano, Professor of Intelligent Systems, and newly appointed Director of the National Centre of Competence in Research (NCCR) in Robotics at EPFL.

Professor Floreano, what is the specificity of your Laboratory, the Laboratory of Intelligent Systems (LIS) at EPFL?

The specificity of my Lab is the origin, evolution, and self-organization of biological systems applied to machines. It is not about mimicking a cognitive system, or imitating human performance. It is really about how very simple components get together and how, with the right simple ingredients, they self-organize and become something that is new and adapted to the environment. I am really interested in principles of biological self-organization and I include here evolution. How did things become so complex and adapted? I am interested in extracting and applying those principles of evolution and self-organization to machines.

For example, my laboratory is very active in the field of Evolutionary Robotics. These are robots where their control system and even their design are not made by us, but the result of an evolutionary process that we let taking place in computers (with physics based simulators) and then put in a machine. If you talk about the origins (of life) and about self-organization, one of the things that always struck me is that nature is capable of doing incredible things in a very small and simple way. Over the years, I have been working on minimalistic solutions for robots and intelligent systems.

In my laboratory, there are people with competencies in software and biology as well as researchers that conceive new mechanical and electronic solutions for robots.

Can you describe one application coming out of your Laboratory?

Most of our work is research-oriented, but applications may develop within the next 15 years. Some of our research avenues however already found applications.

An interesting application is on environmental monitoring, with robots developed for search and rescue. They can actually have a big impact in society. We have for example set a spin-off company, Sense Fly (http://www.sensefly.com), which produces flying platforms for imaging and environmental monitoring. This application raised the interest of agricultural companies for example to monitor large fields, looking at crops, or humidity levels. Here, a personal robot does the job, instead of a helicopter.

“When robots will become successful, they will actually disappear”

Prof Floreano, can you tell us what you think robots will do for us, say in 50 years time?

I would prefer to say: what are the future challenges of society where robots can help? In my opinion, when robots will become successful, they will actually disappear. What I mean is that robots will become successful by melting in everyday objects and in everyday life to the extent that we will not call them robots anymore.

For example, the automatic parking system available in some high-end cars today derive from robotics research of 10 years ago, but we do not call our cars robots. A similar story applies to the robotic components of some automated vacuum cleaners. In the future, many other appliances or daily objects of our environment will become much more autonomous and adaptive, with the purpose of improving the quality of life and of the environment in a natural and sustainable fashion.

This article was written by Lara Rossi, Station-Sud communications Agency