by Ignasi Gómez-Sebastià, Dario García-Gasulla and Sergio Alvarez-Napagao

Assistive technologies are applied to support people in their daily life. Most approaches focus solely on the direct interaction between users – in our case, disabled patients - and the assistive tool, but Artificial Intelligence (AI) has the potential to provide innovative mechanisms and methods capable of taking into account more complex interactions. For instance, such an approach can take into account the important role that third parties may have in user activities, and explicitly reflect the social constraints that apply in the relationship between device and patient. In COAALAS (COmpanion for Ambient Assisted Living on Alive-ShareIt platforms), organizational and normative structures are used to model the sensor network around disabled users as societies, along with the expected behavioural patterns, effectively supporting smart assistive tools that integrate in perfect harmony with the humans around them. The result is an assistive society of ambient-aware assistive tools.

The main goal of COAALAS is to contribute to the state-of-the-art in semi-autonomous and intelligent devices for elderly people. The target population for these supporting devices includes individuals who are independent enough to live autonomously in their community. The role of intelligent devices is to maximize their safety and comfort, thus increasing their quality of life and delaying their institutionalization.

Scientific foundations
COAALAS builds on the results of two European projects: EU-Share-It (FP6-045088) and EU-ALIVE (FP7-215890). COAALAS aims to produce a new generation of AmI devices for elderly people by utilizing several state-of-the-art AI techniques:

  • Autonomy: the device is integrated into the environment and is able to perceive and react in a timely fashion to environmental variables.
  • Proactivity: the device is able to anticipate and take the initiative in order to fulfil its design objectives.
  • Social behaviour: the device is integrated within a community of actors and is aware of social regulations and protocols.
  • Adaptability: the device will modify its behaviour based on the and the other actors around it.

Using a combination of these techniques, COAALAS focuses on making sensors intelligent enough to organize, reorganize and interact with other actors. Sensors have an awareness of their social role in the system -- their commitments and responsibilities -- and are capable of taking over other roles if there are unexpected events or failures. In short: our objective is to create a society of physically organizational-aware sensors able to adapt to a wide range of Ambient Assisted Living situations that could have an impact on the well-being of the user.

Use case
The main purpose of COAALAS is to facilitate necessary and periodic tasks, which in normal conditions would force a physically or psychologically disabled patient to leave their house.

For instance, care-giving tasks involve periodic interaction with a doctor, which might include provision of medication. Tests and validation of a medication coordination system in COAALAS are to be conducted in domotic houses with real users. In this scenario we have identified several actors: the medication dispenser and the rest of the usual domotic house sensors (smoke detectors, domotic doors with identification systems, and so on), the patient, the doctor, the caregiver, and the delivery person. Non-human actors are represented by intelligent agents embedded to the sensors.

The medical dispenser is autonomous and proactive, adapted from previous work done in the EU-Share-It project. The dispenser keeps track of the number of medication doses it has dispensed. Depending on the number of doses left and the medication plan of the patient – ie, the number of doses per day – the dispenser autonomously schedules and sends a request to the patient’s pharmacy for more medication. Therefore, the user does not have to leave his or her house to get more medication.

The dispenser has a social role to fulfil: to provide the correct dose of medication to the user at regular intervals. As the dispenser cannot provide medication once the medication has run out, it will proactively request more medication in order to successfully fulfil its goal. This particular behaviour is the result of the ability to anticipate unusual events and to take action to ensure that goals can still be fulfilled.

In order to achieve a social and adaptive behaviour, we model the social network around the user. It effectively allows for connecting the patient to caregivers and relatives, specifying social regulations for detecting unusual and unwanted situations. Detecting these kinds of events and being able to re-plan the behaviour allow for a fast and adaptive response to potentially harmful situations.

Modelling the society
COAALAS follows the OperA methodology, already used in the EU-ALIVE project, in order to describe organizational models. This methodology enables the identification of the different actors and the relationships among them. For instance, the medical dispenser depends on the delivery person to accomplish medication refills. Furthermore, OperA allows social restrictions and protocols to be defined and be applied to the interactions among actors. For example, the delivery person cannot refill the medical dispenser unless he has checked the medication’s best-before date.

In case an agent is not able to fulfil its obligation, COAALAS provides means for re-organizing the social structure, looking for alternative ways to achieve the goals. For instance, if the medical dispenser is out of medication and the patient needs his dose but the delivery person is unable to arrive on time, a caregiver can be sent to the patient’s house to temporarily fill this role. In case any of the social restrictions is not fulfilled, repair actions are available for returning the scenario to an acceptable state. For instance, if the delivery person has entered the house without identifying himself at the domotic door, he can be sanctioned for this inappropriate behaviour and a caregiver is requested to visit the patient to check that everything is okay.

This project is lead by the KEMLg research group, specializing in intelligent agents - autonomous software entities that observe and act upon the environment, with the aim of achieving particular goals - and argumentation - how to reach mutually acceptable conclusions through premise-driven reasoning.

The ALIVE project is a European project of the 7th framework coordinated by the Universitat Politécnica de Catalunya (UPC) and developed in collaboration with: University of Bath, Trinity College Dublin, Thales Nederland B.V., Tech Media Telecom Factory SL, The University Court of The University of Aberdeen and Universiteit Utrecht.

The SHARE-it project is a European project of the 6th framework coordinated by the Universitat Politécnica de Catalunya (UPC) and developed in collaboration with: Fondazione Santa Lucia, Universidad de Malaga, Telefonica I+D, Universität Bremen, DFKI, Centro Assistenza Domiciliarie Roma.

Links:
https://www.kemlg.upc.edu/
http://www.ist-alive.eu/
http://www.ist-shareit.eu/
http://kemlg.upc.edu/coaalas

Please contact:
Ignasi Gómez-Sebastià, Dario García-Gasulla and Sergio Alvarez-Napagao
Universitat Politècnica de Catalunya, Spain
Tel: + 43 934134011
E-mail: {igomez,dariog,salvarez}@lsi.upc.edu

{jcomments on}
Next issue: January 2019
Special theme:
Transparency in Algorithmic Decision Making
Call for the next issue
Get the latest issue to your desktop
RSS Feed