Getting into Robotics: Davide Scaramuzza
So, what made you decide to pursue academia rather than magic? (Did you always know you wanted to study robotics?
I still haven’t decided which one I want to pursue the most! I am joking 😉 I actually decided that I wanted to become an electrical engineer, when I was 11, after finishing primary school. I was always fascinated by robots and computers. I also always admired people who could repair electronic devices, like my parents’ TV! But the path to become a professional magician and, thus, using magic for living is more difficult, so the decision was easy 😉 I became fascinated by robots when I watched Star Wars as a kid. My father used to tell me bed-time stories about robots, as well.
Somehow, robotics and magic share something in common for me: the goal of magic is to pretend that objects around you are animated and can move, appear or disappear. In robotics, you also make objects move by themselves.
What made you branch out into working with flying robots rather than sticking to ground robots?
Ground robots are constrained to move on the ground and, therefore, their motion is limited to 2 dimensions (or 2.5 half if we want to be accurate). Flying robots instead move in a 3D world, which is extremely helpful in search-and-rescue and remote inspection scenarios, which are the driving motivation of my research and of the NCCR Rescue Challenge. My favourite flying robots are quadrotors, since they can take off and land vertically. The navigational and hovering capabilities of quadrotors make them the ideal platform for exploration, mapping, and monitoring of tasks in search and rescue applications. Quadrotors have the great potential to navigate quickly through unstructured environments, enter and exit buildings through narrow gaps, and fly through collapsed buildings. However, developing flying robots that approach or even outperform the manoeuvrability and speed of birds poses a number of challenges for robotics research in terms of perception, state estimation, planning, and control. It’s the plethora of open challenges (which have been mostly overcome for ground vehicles) that attracted me to flying robots rather than sticking to ground robots alone.
Can you explain your path to professorship?
After finishing my Master’s, I was not sure that I wanted to pursue a PhD. I just knew that I wanted to learn more. I started applying for jobs, but I soon realized that it is not possible to make a change in the technological world if you don’t know more about a topic than what you learn in school. When I first started my PhD (with Roland Siegwart), I wasn’t sure that I really wanted to do it, although after one year it became quite clear that that was the best choice I could have made. At the end of the PhD you are THE expert of your topic.
After the PhD, I got the opportunity to lead the EU project sFly, as a postdoc. That project was my first experience as a project coordinator, and I soon realized that I enjoyed the possibility of deciding research directions and, thus, leaving my signature in something good and useful for humanity. After my postdoc, I got a job offer at NASA/JPL. That’s the moment that I realized I needed to decide the job I really wanted to do in the future. It was not an easy choice, but I realized that I wanted to be a free thinker rather than be constrained to a project and that pursuing an academic career would give me the freedom to do the research I want to do! So, instead of accepting the job at NASA, I started my academic career. I have stayed and, so far, I have never regretted it!
What advice for those who would like to consider studying robotics in the future? ..and to those who wish to follow into academia?
My advice to those who are considering studying mobile robotics is to study well electrical and mechanical engineering and, especially, computer science. Most of the advances made in robotics over the last decade have been done in computer science, and especially in perception and machine learning.
My advice to those who are starting or have already started their PhD is to read a lot. But also practice sports and cultivate your hobbies in order to free your mind and facilitate generation of ideas! Read, read and never stop reading scientific papers. Only by learning from others can you get the best ideas. When you approach a new problem, it is good to start thinking and developing your own solution. But I often see PhD students pursuing their own solution but waiting till the last days before a conference deadline to do a literature research and eventually find out that others had already thought of a similar solution. Sport and hobbies are very important components for good research. I had most my ideas not at the work desk, but while doing completely different things outside of work!
For those who are finishing their PhD and want to pursue academic career, I recommend to move to a different lab in order to broaden their ideas. Sticking to the same lab doesn’t expose you to different people, research fields, and cultural environments. The fact of living in a different environment and, especially, in a new country, is definitely a plus. It faces you with new challenges in life (like learning a new language) which can make your passion for research grow even stronger.
Finally, if you want to become a professor, another important thing is: become known by people. Go to conferences, do networking, and present your staff in other labs, departments, universities that you would like to join one day. Many of those in professor committees most likely don’t read as many papers as they did when they were younger. So, the only way for them to get to know your work is if you show it to them directly!