Interview with Antonio Andriella
Research Scientist at PAL Robotics, Antonio Andriella recently won the prestigious Georges Giralt PhD Award for his thesis on personalising robotics assistance to help dementia patients, and is currently focussed on the PRO-CARED project to help future generations in education and learning through robotics. We had the chance to interview Antonio about his PhD thesis research and award, and his work on project PRO-CARED to offer robotics support in education.
Antonio’s main research interests are in the areas of human-robot interaction (HRI) and human-centered design technologies. He has organised workshops on topics related to trust, AI, ethics and personalisation at robotics conferences, as well as serving as guest editor of several special issues of journals such as the International Journal of Social Robotics, and Paladyn Journal of Behaviour and Interaction Studies, and as associate editor at IROS and RO-MAN.
Georges Giralt PhD Award by EU Robotics and cognitive training therapy to patients with dementia
Antonio Andriella was presented with the best thesis award for application in AI by the Spanish National Research Council – the CSIC for his doctoral thesis work “Personalizing robot assistance for cognitive training therapy” directed by Dr. Guillem Alenyà and Prof. Carme Torras of the Institute of Robotics and Industrial Informatics (IRI, CSIC UPC). This was followed by the prestigious Georges Giralt PhD Award for the best PhD thesis in Europe – by EU Robotics, presented to Antonio at the ERF 2023 (European Robotics Forum).
Regarding the research for his PhD Antonio told us, “It’s focussed on developing new methods to enable robots to autonomously learn to exhibit personalised behaviour. I explored this in the context of building robots that provide cognitive training to people affected by cognitive impairments, in particular those with dementia.”
Antonio continued, “I worked with PAL Robotics’ TIAGo robot and replaced the robot head with a screen with an expressive face. The robot has some human-like characteristics such as personality, communication style and backchannelling in order to make its behaviour more understandable. The robot is capable of delivering cognitive training therapy to patients with dementia by providing them with the degree of assistance that best suits the individual’s specific needs.”
Explaining further he told us, “the robot autonomously provides a test to the patient- they have to sort objects or tokens according to certain criteria. The robot monitors whether they are doing the test properly, and if not the robot provides help. In order to do this, first of all we monitored and learnt how human therapists do the work, then we map that on the robot.”
“The robot’s functionalities were first tested with a healthy population (more than 600 people) and finally evaluated with 22 patients at Ace Alzheimer Center Barcelona. The overall results were promising: the robot succeeded in correctly administering the exercises to the patients by tailoring assistance to them and to the extent to be as effective as the human therapist.
He added, “however, as roboticists, we need to be realistic. We don’t want to replace therapists, we want to enhance their capabilities, yet we have an increasing number of adults that need help. Indeed, robots can be very useful tools to address this societal gap, especially in repetitive tasks such as cognitive training therapy.”
“In the future, we could envisage a scenario in which a therapist could set up the initial robots’ behaviour and then the robots can provide cognitive exercises to several patients at the same time, freeing up the workers’ time to spend their time on activities that better leverage their skills.”
VIDEO: animation of thesis “Personalizing robot assistance for cognitive training therapy”
There are 50 million people worldwide living with dementia, and this number is projected to triple by 2050. Behavioural therapy such as cognitive training can maintain or improve the patients’ general cognitive functioning, eventually prolonging ageing in place. However, the number of patients who require such treatment greatly exceeds the number of healthcare professionals’ availability. Robots have the potential to bridge this gap by enhancing therapists’ effectiveness in delivering cognitive therapy. Despite the recent advances, many questions still need to be addressed regarding the design, development, and deployment of social robots for such purposes. Read Antonio’s full thesis here.
VIDEO: Study with the Robot Therapist to Evaluate CARESSER
PRO-CARED project: pro-active behavior and user interaction in education
Antonio also works on the EU-funded PRO-CARED project – he told us “during my PhD research, the idea was to design a set of levels of assistance that the robot could provide and which could best meet the needs of the user. However, we didn’t always know when the time was to provide assistance. The PRO-CARED project is focussed on robot proactive behaviour and we aim to learn when it is time to interact with the user, what assistance and when to offer it – this is very important especially for long term interaction. The robot needs to understand when to approach the human. I’m particularly interested in proactive behavior, and also mental models – such as the theory of the mind (connected to the capacity to understand other people by ascribing mental states to them).”
Antonio continued, “I spent one year in Naples – with Prof. Silvia Rossi in the PRISCA (Intelligent Robotics and Advanced Cognitive System Projects) lab – working on algorithms to enable robots to be proactive. We ran four experiments, some of the results of these are already published and for others we are currently preparing papers to send to journals.”
“The experiments were connected to proactive behaviour in sequential memory tasks and adaptive robot personality in conversational tasks. Overall, the proactive behaviour of the robot was perceived by the participants and they could improve their performance because of this during memory tasks.”
“As part of the PRO-CARED project the robot will help students – as a tool for teachers to provide specific or personalised lessons – this may be in a high school or even in a language school, to help teach the students.”
In the PRO-CARED project, PAL Robotics is involved with Antonio Andriella to develop a new enabling technology: personalisation. Personal robots are ready to create a unique, personally tailored educational experience and can provide a powerful tool for teachers to enhance their lessons. PRO-CARED is designing and developing mechanisms for personal robots to proactively interact with students, providing them with adaptive and personalised assistance in an educational context.
Future work: deploying ARI in education to interact with students
Regarding what’s next in his work, Antonio told us, “at the RO-MAN Conference in August I will present my latest work and a session on robot personalisation. This includes co-organising the workshop: WARN: Weighting the benefits of Autonomous Robot persoNalization, and chairing the special session: Short- and Long-Term Personalisation in Social HRI. I’m also co-author of the article: Sweet Robot O’Mine – How a Cheerful Robot Boosts Users’ Performance in a Game Scenario with F. Vigni and S. Rossi which will be presented at RO-MAN.”
Antonio continued, “later we’ll be working on integrating PAL Robotics’ ARI robot in PRO-CARED for education. The idea is to deploy the robot for a month – to see how the robot interacts with students without us there to provide assistance. We’ll be gathering a lot of feedback and the hope is to develop a robot that can help teachers and improve the skills and the performance of the students, as well as pushing the limits of robotics.”
Challenges and benefits: robots as a tool to help in education and healthcare
When asked about the challenges of robotics in education, Antonio told us, “the main challenges are unpredictability in human behaviour. Also, we want to have a robust robot architecture that can interact in a deployment of one month, as we don’t want to avoid the need to intervene regularly.”
Antonio added, “Another challenge is – why are we using a robot and not a tablet? Previous work tells us that usually people prefer to interact with a robot as it’s more engaging. However, we know very little about what happens after the “novelty” effect wears off.
With this project, we aim to understand how people react and interact with the robot after their first encounter. Very little work has focussed on deploying robots in real-world scenarios, this is why we decided to evaluate robots with real end-users in educational contexts.”
Lastly, Antonio talked of the ability of robotics to help in healthcare and education, telling us, “there are some societal gaps, for example, an increasing number of people that need healthcare services, and in education big classes where there are a lot of students. Teachers need to teach to the average level, and can’t always help every single student. Robots can help with this specific need, as a tool for the teacher to be able to provide personalised lessons to students, especially those that need more help. In healthcare alike, our research has shown that robots are suitable tools to provide personalised assistance, in order to help both patients and healthcare professionals.”
Antonio Andriella: background and research interests
Antonio is a Research Scientist at PAL Robotics. He received his PhD with a thesis entitled “Personalising robot assistance for cognitive training therapy” from the IRI under the supervision of Dr. Guillem Alenyà and Prof. Carme Torras. He was an Early Stage Researcher in the Marie Skłodowska-Curie Innovative Training Network (ITN) project called SOCRATES, which focused on Quality of Interaction in care for older people. Prior to IRI, he worked at Cogisen as Artificial Intelligence lead. He was a visiting student at the University of Naples (2021) and the University of Hamburg (2018-19), as well as a visiting researcher at the University of Naples Federico II under the supervision of Prof. Silvia Rossi.
We would like to thank Antonio Andriella for taking the time to talk with us. You can learn more about him and his previous work here. If you have any questions for Antonio or the team at PAL Robotics don’t hesitate to get in touch with us. Visit our website to find out more about our social robots, including humanoid social robot ARI, and humanoid mobile manipulator robot TIAGo. For more articles and interviews, check out our blog on robotics and technology.
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under Marie Skłodowska-Curie grant agreement No. 801342 (Tecniospring INDUSTRY) and the Government of Catalonia’s Agency for Business Competitiveness (ACCIÓ).