The THOMAS architecture in Home Care scenarios: A case study

https://doi.org/10.1016/j.eswa.2009.11.017Get rights and content

Abstract

Today, the need for architectures and computational models for large scale open multi-agent systems is considered a key issue for the success of agent technology in real world scenarios. The main goal of this paper is to describe a case study in Home Care scenarios applying an abstract architecture and a computational model for large scale open multi-agent systems based on a service-oriented approach. The architecture used is THOMAS, which specifically addresses the design of Home Care systems. This paper presents services examples for the management of a dependent home environment, and demonstrates the new features of the proposal.

Introduction

Home Care is one of the objectives of pervasive computing, and dependent people require new solutions that can take advantage of technological advances which provide novel and fundamental services (Angulo & Tellez, 2004). The vision of pervasive computing is to improve the quality, access, equity and continuity of health care (Angulo & Tellez, 2004). In this sense, intelligent environments can improve health care services and can have a high social impact, especially in Home Care services for chronically dependent patients (Augusto & McCullagh, 2007). Because Home Care requires effective communication as well as distributed problem solving (Augusto & McCullagh, 2007), multi-agent systems can facilitate the development of pervasive Home Care environments. Moreover, agent-oriented methodologies provide mechanisms for modelling distributed, inter-operable and secure systems by taking social and organizational considerations into account. Agents are autonomous software entities (Camarinha-Matos & Afsarmanesh, 2004) able to interact with their surroundings. They are highly capable of adapting to changes, and can allow for integration with multiple devices, sensors and humans.

The continuous growth of the dependent people sector has dramatically increased the need for new Home Care solutions (Anastasopoulos et al., 2005, Corchado and Laza, 2003). Furthermore, commitments to meet the needs of this sector suggest that the current systems are in need of modernization. Multi-agent systems (Want, Pering, Borriello, & Farkas, 2002) and intelligent device-based architectures have been recently explored as supervisor systems in health care scenarios (Angulo & Tellez, 2004) for elderly people and for Alzheimer patients (Corchado & Laza, 2003). These systems are capable of providing constant care in the daily life of dependent patients (Carrascosa, Bajo, Julian, Corchado, & Botti, 2008), predicting potentially dangerous situations, and facilitating a cognitive and physical support for the dependent patient (Augusto & McCullagh, 2007). Taking these solutions into account, it is possible to assume that multi-agent systems can further facilitate the design and development of pervasive environments (Corchado, Bajo, de Paz, & Tapia, 2008) and improve the services currently available by incorporating new functionalities. Multi-agent systems add a high level of abstraction with respect to the traditional distributed computing solutions. They also facilitate the analysis and design of the problem in terms of artificial intelligence systems. This allows a greater flexibility for incorporating human behaviours into the agent’s structure. Multi-agent system technology makes it possible to cover a broad area of problems. Typical problems are systems in which there are several entities (Requesters) which may require one or more elements or services from other different entities (Bidders). In the area of Home Care, for example, Requesters would be patients and Bidders would be companies which provide services, such as identification, localization, home automation services, or warnings and alerts. Obviously, the development of these types of systems is complex and, therefore, it is necessary to analyze the intrinsic characteristics of these typical application environments in detail.

The aim of this research project is to present a case study applying the THOMAS (MeTHods, Techniques and Tools for Open Multi-Agent Systems) multi-agent architecture. THOMAS has been used to develop a case study for supervising and monitoring dependent patients at home. This multi-agent system offers a series of functionalities including an automatic reasoning and planning mechanism for scheduling the medical staff working day, an alert system, a location and tracking system, and an identification system. The medical staff has been provided with PDAs and mobile phones, as well as with Java Card tags, and the home environments have been equipped with presence detection sensors, access control mechanisms, door opening devices and video cameras. The multi-agent system monitors the daily routine of the patient and detects dangerous situations. If any anomalous situation is detected, the alert system is used to obtain medical assistance.

The remainder of this paper is structured as follows: Section 2 provides an analysis of related studies; Section 3 presents the proposed architecture model as well as a description of the services offered by each one of the modules that make up the reference model; Section 4 shows an example of an implementation, highlighting the new possibilities provided by this type of architecture, and presents a specific approach for Home Care management; finally, some conclusions of this study are shown in Section 5.

Section snippets

Related works

Dependence is a permanent situation in which a person needs important assistance from others in order to perform basic daily life activities such as essential mobility, object and people recognition, and domestic tasks (Costa-Font & Patxot, 2005). There is an ever growing need to supply constant care and support to the disabled and elderly, and the drive to find more effective ways of providing such care has become a major challenge for the scientific community (Nealon & Moreno, 2003). The

THOMAS architecture model

THOMAS architecture basically consists of a set of modular services. Though THOMAS feeds initially on the FIPA architecture, it expands its capabilities to deal with organizations and to boost its services abilities. In this way, a new module in charge of managing organizations is introduced into the architecture, along with a redefinition of the FIPA Directory Facilitator that is able to deal with services in a more elaborate way, following Service Oriented Architectures guidelines. As

Applying THOMAS to Home Care

Home Care facilitates the interconnection between dependent people and their environment and medical staff (doctors, nurses and personal assistant), delimiting services that each one can request or offer. The system controls which services must be provided by each agent. The internal functionality of these services is the responsibility of provider agents. However, the system imposes some restrictions regarding service profiles, service request orders and service results.

A description of the

Results and conclusions

An important issue in the development of real open multi-agent systems is to provide developers with methods, tools and appropriate architectures which support all of the requirements of these kinds of systems. This paper has studied this problem by proposing a case study on abstract architecture for the development of virtual organizations. Moreover, the proposal aims to instigate the total integration of two promising technologies, multi-agent systems and service-oriented computing, as the

Acknowledgements

This work was supported by the Spanish Ministry of Science and Technology project TIN2006-14630-C03-03 and the JCyL SA071A08 Project.

References (22)

  • Cesta, A., Bahadori, S., Cortellesa, G., Grisetti, G., & Giuliani, M. (2003). The RoboCare project, cognitive systems...
  • Cited by (34)

    • Enhancing smart-home environments using Magentix2

      2017, Journal of Applied Logic
      Citation Excerpt :

      However, as in the majority of approaches, the user agents are assumed that all of them play the same role. Bajo et al. [3,16] present a previous version of the THOMAS architecture applied to a similar scenario used in this paper. Among existing implementation platforms of multi-agent systems, JADE is the most popular.

    • Sensor-driven agenda for intelligent home care of the elderly

      2012, Expert Systems with Applications
      Citation Excerpt :

      This status means actions such as standing, sitting or lying (with a very different impact whether on a bed or on the floor). In Bajo, Fraile, Pérez-Lancho, and Corchado (2010), a set of modular services for supervising and monitoring dependent patients living at home is presented. A multi-agent system offers a series of functionalities, including an automatic reasoning and planning mechanism for scheduling the medical staff working day, an alert system, a location and tracking system, and an identification system.

    • Literature review of managerial levers in primary care

      2020, Journal of Health Organization and Management
    • A Market-Based Scheduling Mechanism Design for Cost Reduction in Home Health Care

      2019, Journal of Integrated Design and Process Science
    View all citing articles on Scopus
    View full text