Aye Robot ... how Scotland is creating the world’s most socially aware droid assistant
LOST in a shopping centre? Unsure of your train platform? Want a picture in an art gallery explained? The information may soon be provided by robots, developed in Scotland, which not only give you the facts but second-guess that you need help by reading your body language and subtle social cues that betray what you are really thinking.
Sounds far-fetched? Not according to a team of robotic experts from Glasgow University working to create the world’s most “socially aware” robot. Using existing technology they are creating a machine that can “spontaneously interact” with strangers in public spaces such as shopping centres, airports, stations and museums.
By adapting the diminutive, £24,000 Pepper robot – first marketed in Japan and now in the US as an “endearing companion” – Scottish computer scientists aim to take the technology to a new level by programming it to interact empathetically with passers-by.
The main goal of the four-year-long MuMMER (MultiModal Mall Entertainment Robot) project – which also involves Pepper creator SoftBank Robotics, researchers from Heriot-Watt University and the Idiap Research Institute in Switzerland – is a socially intelligent robot that will be tested in Finland’s Ideapark shopping centre.
This “entertaining and engaging robot” will help lost or confused shoppers and give out information on deals or discount vouchers.
The team claims the robot will have a wide range of other applications in public spaces and could even be used in hospitals and other healthcare settings to provide everything from entertainment in waiting rooms to triage services.
Dr Mary Ellen Foster, Glasgow University lecturer and co-ordinator of the MuMMER project, said: “Pepper is a robot that is already going into public spaces and attracting a lot of attention, but its traits are not very interactive. What it does is scripted. It’s operating more like a touch screen.
“We are working on a robot that is able to have a conversation and make judgments about who it can approach and what sort of approach it should make. In Glasgow, we are looking at social signal processing and how the robot can understand those signals.
“For example, how can a robot tell if we want to talk or if we might be scared or avoiding it? The robot could respond by being more polite, by trying to be more engaging and entertaining, or more calm.”
Already Pepper, which stands at just under five feet tall, can recognise the human voice in 20 languages and detect if an adult or child is talking. Earlier this year, the robot was enrolled in classes at Shoshi High School in Waseda, in Fukushima, Japan – mak- ing it the first robot to “study” alongside human students, demonstrating its ability to learn. The enhanced Peppers will be tested in the lab as well as the university’s public spaces, including the Hunterian Museum and Art Gallery, before going to Finland and other partner countries for cross-cultural comparisons. Team member Professor Alessandro Vinciarelli, also from Glasgow University, said a key challenge was creating robots with the technical capability to read human emotions in noisy and crowded public spaces such as shopping malls. Creating sensors, microphones and cameras capable of picking up non-verbal signals is one of the issues that the project will grapple with. “We are looking at creat- ing a socially intelligent machine that can understand very subtle signs,” he said. Recent work in the development of robots has included creating machines capable of working with autistic children as well as companions, such as Paro, an interactive furry seal robot developed in Japan, which aims to replicate pet therapy.
So far, all applications have been experimental though Vinciarelli claims they could be mainstream within a decade.
But he admitted there were some risks associated with the development of robots including over-reliance and threats to privacy. “Robots should work alongside humans to support them, not to replace them,” he said. “And when machines can understand our emotions and are in our public spaces the entire concept of privacy needs to be redefined.”
We are looking at creating a socially intelligent machine that can understand very subtle signs
Can I help you? Pepper the robot is coming to a public place near you soon Photograph: Getty