AI and Personal Autonomy

 


We are going to talk about Artificial Intelligence. However we will not introduce any anthropomorphic android to you. Instead, we will talk about a practical project with a huge social impact. Professor Ulises Cortés, and his PhD student Cristian Barrué with the help of other members of the KEMLG research group are working on the development of technological support system that is meant to help senior and disabled people in their daily life.

The project uses intelligent agents and machine learning techniques embedded in different tools and devices in order to make the life of disabled people and their caretakers easier. In particular, several techniques have been developed in order to learn different paths within a house or to predict the needs of an impaired person according to past patterns of behaviour. Several agents acting on the motor and direction controls of an electrically powered wheelchair learn how to steer it towards a possible final destination within the home without the person having to intervene. By applying Case-Based Reasoning (Case-Based Reasoning, a method that learns by analogy from past experiences), intelligent agents can help someone follow safely a path. They can also help in the situation, very common in old age, when one remembers about something but has no idea about where it could be. The systems learns frequent associations between objects (like tablets) and specific situations and contexts. From this knowledge, it is able to guide the wheelchair towards the place where the object should be. With this kind of help, the feeling of autonomy of impaired people will be extraordinarily enhanced and their quality of life much improved.

This is the KELMG contribution to EU Sixth Framer work Programme Project "SHARE-it". This project has wider goals have to do with how disabled and old people interact with technological systems in general.

The main goals of the project are:

• To explore the benefits of artificial intelligence in order to build elements that increases the autonomy of their users.
• To research and introduce new forms of shared autonomy.
• To develop technological solutions that once combined with other technologies will give user better mobility.
• To build multimodal and adaptable interfaces for these type of patients.
• To help the healthcare services in a more efficient way.
  

 

KEMLG is working specifically in applied agent technology and machine learning methods for wheelchairs and walking aids. They have worked in several prototypes starting with the semi-automatic wheelchair Rolland III and moving on to the semiautomatic platform Spherik. This platform is used as an ancillary mobility platform that can hold a patient traditional wheelchair and act as its moving engine and assistant. It uses ball-shaped wheels, hence its name. The combination of technologies used in Rolland and Spherik is being applied now to the Intelligent Walking Aid.

With these platforms it is possible to detect the location of the patient within his or her home. It is also possible to fine tune and adapt automatically the platform to the needs for the user. At the same time, all the health constants of the patient are under continuous monitoring and sent to the health support professionals.

It is expected that the project will yield further knowledge and solutions in aspects that go beyond its application to health. That is, to areas of research within Artificial Intelligence and Machine Learning. For example, there is an important contribution in Knowledge Representation and Acquisition as well as to high-level reasoning and goal-directed behaviour in real world situations. It will also provide a general intelligent infrastructure based on agent technology that will be a common resource for decision making, reasoning, learning, communication and other services that could be built upon it. The result will be encapsulated in a set of generic software components that are meant to be of high quality, scalable and robust.

SHARE-it partners are: Fondazione IRCCS Santa Lucia (IT), Universidad de Málaga (ES), University of Bremen (GE), Telefónica I+D (ES), Ana Aslan International Foundation (RO), DFKI Bremen (GE), Centro Assistenza Domiciliare Azienda Sanitaria Locale RM B (IT).

The project has been running for eight months now for a total duration of two years. We wish the best luck to the KEMLG team in attaining a much needed social improvement.
 

 

- Cristian Barrué, Ulises Cortés, Roberta Annicchiarico. Shared Autonomy in Assistive Technologies. IWANN 2007: 1067-1073.

- FP6-IST-045088 Project SAHRE-IT. Supported Human Autonomy for Recovery and Enhancement of cognitive and motor abilities using information technologies (SHARE-it): http://www.ist-shareit.eu/shareit

 

ilapuente@lsi.upc.edu

(Back to the Newsletter)