Although this project is using humanoid robots, I don't see any reason why the AI couldn't be used in a PLEO. There might be a need to have PLEOs at different prices, depending on their capabilities, but I, for one, would certainly be willing to pay whatever it costs. As for rehabilitation purposes, I think kids in general would respond better to a PLEO, than a humanoid robot.
I hope Innvo Labs can find a way to get on board this project.
Total recall - Plymouth leads European project into robot memoryDate: 27 August 2010
SummaryA multi-million pound project has begun to design a new breed of robot that can form memories and engage in social interaction.
Further detailThe European Commission-funded work is hoping to overcome traditional limitations in artificial intelligence by building social cognition for the robots. To achieve this the robots will be connected to internet-based programs with much greater computational power.
The project, codenamed ALIZ-E, will involve eight universities and institutions across the continent, as well as a hospital in Milan which is hoping to use the robots in the rehabilitation of young patients.
The University of Plymouth is leading the four-and-a-half-year, Euro 8.3m project, and will be focusing on those aspects relating to memory. Project head Dr Tony Belpaeme said: “The aim of the ALIZ-E project is to explore how human-robot interactions can be extended from minutes to the scale of days, thus forging longer-term constructive bonds between robot and user. The ALIZ-E project will use the principles of embodied cognitive robotics to create agents capable of sustaining believable, in-depth social relationships with young users, over an extended, potentially discontinuous time frame.”
Currently one important limitation of robots is the size of their on-board processors, which are incapable of forming ‘memories’ over an extended period of time. But by using experts across Europe, the team is hoping to create a system which will enable the robots to store and recall information and modify their behaviour on the basis of previous experiences.
The partners have all been selected for specific roles, including speech recognition at the National Research Council in Italy, recognition of emotion at the University of Hertfordshire, and development of the ‘cloud’ computing techniques that will be at the heart of this memory system, in France.
Dr Belpaeme said the ALIZ-E program would look to develop a viable alternative to animal assisted therapy (AAT) – a practice whereby children in hospital are given animal companions during serious, long-term illnesses. He said: “We will specifically explore robot-child interaction capitalizing on children’s open and imaginative responses to artificial ‘creatures’. When a child is in hospital and has to learn how to manage a long-term condition such as diabetes, it becomes very important to develop effective communication.
“The theory is that the robot acts as a companion and a communication channel between patient, parents and hospital staff. We will essentially be taking robots out of the lab and putting them to the test in a health education role, with young diabetic patients, in a busy pediatric department at the San Raffaele hospital in Milan.”
There are currently 20 of the robots being worked on across the partnership. Each weighs 4.3 kilos and stands 58 centimetres tall.
The project team includes:
* Vrije Universiteit Brussel (Belgium)
* Deutsches Forschungzentrum für Künstliche Intelligenz (Germany)
* Imperial College London
* The University of Hertfordshire
* National Research Council - Padova (Italy)
* The Netherlands Organization for Applied Scientific Research
* GOSTAI (France)
* Fondazione Centro San Raffaele del Monte Tabor (Italy)
More info at:
http://www.scientificamerican.com/gallery_directory.cfm?photo_id=D2AA87DD-0F13-F40C-C50D2364409B95A1&sc=DD_20100902 