Born – Malta
PhD – The University of Manchester, UK
Joined University of Liverpool – 2021
Position – Senior Postdoctoral Researcher, Department of Chemistry
Group Name – Cooper Group
What is your research about?
My research sits at the intersection of robotics and machine learning. I study how we can use robotic manipulators for different real-world applications. I am particularly interested in working on robotics for applications where it would not be safe for humans to go, such as my previous project within the UKRI National Centre for Nuclear robotics, or where it is repetitive and tedious for humans, such as my current project on using robots to help chemists with mundane tasks in material discovery experiments.
The questions that my research addresses centre around making robots safer and smarter in real-world deployment. I am truly excited about robots making a positive impact on humanity, especially in situations where it could be dangerous or hazardous for humans to be. Within my current role, the scope of my research is to create intelligent ‘robotic scientists’ which are capable of robustly carrying out closed-loop experiments. In simpler terms, I am exploring how following human hypothesis, an experiment is carried out using these robotic agents with minimal human intervention, and the results produced are fed back to create new experiments leading to accelerated discovery of new materials.
What or who first inspired you to be interested in your research subject?
There were different people and occasions that led to me pursue my current research subject. In high school, maths was my favourite subject due to the teachers who instilled my passion for the subject. From then on, I always wanted to choose a career that would allow me to pursue this topic in an applied way.
During my undergraduate studies at the University of Malta, we built a combat robot for an IEEE student robot wars edition, which was when I started becoming more interested in robotics. My master’s supervisor (and then postdoctoral supervisor), Prof. Michael Mistry showed me how to adopt methods from human sensorimotor control to improve robotic manipulation. From then on, I wanted to continue working in research projects that would allow me to pursue this research avenue, which resulted in me pursuing a PhD in cognitive robotics.
Thanks to my PhD advisor, Prof. Angelo Cangelosi, who gave me the support and freedom to pursue this topic, I knew that continuing research in robotics is what I was passionate about. As a result of the COVID-19 pandemic and labs being shut down, I was itching to go back to the lab and hence, the opportunity at the University of Liverpool working on intelligent robotic manipulation for material discovery was the ideal place for me to bring my expertise to a new, yet very important, field.
What are you most proud of achieving during your research career so far?
Coming from a really small island in the Mediterranean Sea where robotics is still at its infancy in terms of career opportunities, I feel very lucky and proud to be able to pursue a research career in this field. However, to be more specific, I think having been awarded a Marie Skłodowska-Curie Actions doctoral training scholarship which not only gave me the educational and financial support throughout the duration of my PhD, but also the opportunity to carry out secondments at the University of Edinburgh and Italian Institute of Technology, would probably be close to the top of my achievements so far.
What techniques and equipment do you use to conduct your research?
My research relies on GPU-accelerated hardware for training and deploying machine learning models and robotic platforms, mostly robotic manipulators.
Which other subjects are important for your research?
Apart from computer science – in terms of algorithms, programming, and software engineering skills, and engineering, control and electronics – maths is also essential, and in particular probability theory and calculus. As robot learning is also inspired by human motor control and human cognition, I believe cognitive science and psychology are also important. Within my current role, chemistry has been quite fundamental due to the robot deployment being carried out in a materials discovery laboratory.
What is the key to running a successful research group?
I believe that there are two key criteria: focusing on ways how to bring the best out of each member and building a group with complementary skills. First and foremost, it is important to understand what the personal goals of researchers are and ensure that their work helps them move closer towards their goals. Having researchers excited and motivated is vital to ensure that the group’s research is pushed forward.
Secondly, I believe that a multidisciplinary group is essential to solve real-world challenges and hence the group should be made of people with different backgrounds. Currently as part of the ERC ADAM project where I am the lead postdoctoral researcher at Liverpool, the group is composed of different researchers, from experimental to computational chemists, engineers to computer scientists, and this is essential to complete the project.
What impact is your research having outside of academia?
In my current project (ERC ADAM), we are focusing on using robotics for materials discovery, specifically for closed-loop experiments. In the past couple of years, scientists were able to develop, test and produce at large-scale a new vaccine for the COVID-19 virus within a year. Using robotics and automation, this has the potential to be accelerated even further. Hence, the core idea of what we are currently doing at Liverpool is radical and at the forefront of laboratory automation: smart, ‘robotic chemists’ that can carry our experiments without any human input. Success here could have a huge impact on progress in materials discovery and in topical sectors such as the pharmaceutical industry. It is an extremely exciting research area and there is real scope for applying robotics here to accelerate research with huge societal impact.
How do you plan to develop your research in the future?
Currently, I am pursuing research endeavours in robot learning to create the next-generation of ‘robotic scientists’, that are able to generalise to new experiments, have the ability to analyse experiments and recover from errors using multimodal sensory feedback and exhibit skill-set that is representative of a human chemist. With my research experience and current role, I hope to continue pursuing research in this exciting direction. Thanks to my current supervisor, Prof. Andy Cooper, and the support of the Leverhulme Research Centre for Functional Materials and the European Research Council, I have already started working on some of these problems but there is still a lot more that needs to be done.
What problem would you like to solve in the next ten years through your research?
Probably, as any other roboticist, my dream is to have robots that can generalise across different tasks and environments in the real world. Particularly, in my case, I hope that our ‘robotic scientist’ would be deployed in real-world laboratories to accelerate academic and industrial research. In the next 10 years, I hope to be able to push robot learning to the forefront of laboratory automation, and move towards having collaborative robots working with and close to humans.
What advice would you give to someone considering a career in research?
The main advice would be to be passionate about the topic you choose to pursue and persevere – I think both are important characteristics to have as a researcher. Most ideas fail during the first trial, and even if you are passionate about it, it can be discouraging to see this. Hence it is important that the researcher has the strength to persevere and keep trying. With this in mind, a career in research can be very rewarding as you are continuously pushing the boundaries of what has been achieved, and in the field of robotics, this can be done in both academia and industry.
Where can readers learn more about your research?
You can find out more about my research by visiting https://gabriellapizzuto.github.io
Back to: Faculty of Science and Engineering