Robotics is his job, the research and development laboratory in Engineering his “kingdom”. Philip Wright, a mechanical engineer, has been working in the field of research for six years, dealing with machines and the algorithms which manage them.
“The goal of those who work with robots – explains Philip immediately – is not so much that of imagining a figure that can replace a human being, but one which is able to support him, to help him in performing repetitive, heavy tasks, that do not require special skills”. This is the case of the humanoid robot Pepper, on which Engineering and the research group of which Philip is a member have been working for over a year and which will be able to support humans in taking care of elderly people. “The project began by integrating a motorized medical walker, already designed and built by Engineering, with Pepper the robot. These, together with other sensors useful for monitoring the state of health of the patient, have the objective of delaying hospitalization for the elderly, with all the consequent benefits for patients, their families, but also for the community. Pepper will not be able to replace caregivers, but it will be able to help the elderly, for example, to do some daily exercises useful for rehabilitation, or it will be capable of interacting with him/her by reminding that some activities need to be carried out. Much is being done to improve human interaction, as Pepper will need to stimulate people and win their trust. A difficult thing, but certainly not impossible”.
The project, which will start its testing phase with a group of patients in 2020, is one of the projects Philip cares most about, as he has been working on it since the very first conception phase. “With Pepper I had the opportunity to make the most of my mechatronic skills in a project team with experts in biomedicine, cognitive and social psychology and, of course, information technology”.
How is it possible to teach a humanoid to interact with humans?
“Human-machine interaction is one of the aspects we are currently studying in the laboratory. Our goal is to make the interaction between Pepper and the patient as complete, simple, but effective as possible. Many technologies on offer today make this increasingly possible, but to create a single effective and robust system, which possesses the language and understanding of humans is not yet possible. What we can do is to start working on the context in order to limit the interaction and to be able to always redirect it towards what we know, bringing the subject back to known questions and making sure that the robot always, or as much as possible, understands what is being said. In addition to understanding and interacting, another interesting aspect is that Pepper can act on its own initiative, for example by starting a conversation. For example, if it detects a long period of silence, after a while it may be the first to interact by asking a question. This can be very helpful, for example, in the health field to stimulate patients: it can help them to improve their memory, particularly for those who begin to lose this over the years, or for people with particular disorders which require periodic speech or cognitive exercises. A machine is certainly able to be methodical in its updates, if it is also capable of stimulation, with the right approach, it can be a fundamental tool. A system like this one can have various applications in many other scenarios, and can be a useful tool for our daily lives”.
How far are we from human-machine interactions which can replace people-to-people interactions?
“We are not too far away if we imagine a situation where a human being is being helped and not replaced. Interaction has improved a lot; technological progress supports the solutions we are putting in place, also in terms of understanding the natural human language. We have various reusable material at our disposal, so we always start from something made by others but that can be improved. Then, in research, anything new that is designed and implemented is useful for the work of others and this is a nice plus in relation to motivational levers”.
How does the typical day of a Robotics Researcher unfold?
“Days are never the same, and this is one of the positive aspects of this job. Let’s say that the basic activities are: design, software development and testing, but a lot of the time is also spent on alignment and coordination calls with project partners, and on documenting what is done”.
What is the most challenging aspect of this job?
“One of the things I like most is to look for the limits of technology and to test myself in order to overcome them. It’s very stimulating, but it’s easy to get lost in this. Research time for projects is limited and you constantly need to bear this in mind and take it into account. Another difficult aspect is the need to coordinate with other researchers of the partners, to keep a balance and manage relationships which are almost always in English, a language which is not ours. The positive thing is that research environments are often dynamic, young and we Italians, despite some stereotypes that we are burdened with, are perceived by European colleagues as versatile, creative and dynamic. Sometimes, we can get a bit lost in this ability to be flexible, but there is always a member of the team that leads us back onto the right track”.
What is the best course of study for this profession?
“I would say that a degree is indispensable, and based on my experience, engineering or computer science would be preferable. In addition to this, as in all professions related to IT, you must be willing to never stop studying, reading, learning from others”.
Advice on books, newsletters, blogs to follow?
“As for books, I would suggest Life 3.0. Being human in the age of artificial intelligence by Max Tegmark, whilst for the newsletters to follow I would say the ones by Cordis Europa, which is useful to keep up with new research projects, and by Robohub which gives all the news on robotics. Other useful tools are also wiki.ros.org, a robotics community, and www.ieee.org“.
