Service tailoring: a method and tool for user-centric creation of integrated IT-based homecare services to support independent living of elderly

This thesis addresses the problem of supporting independent living of elderly people through IT-based homecare services. Independent living is seen as one way to deal with the consequences of an aging population (especially in industrialized countries), which include rising healthcare expenditures and a required shortage of healthcare professionals. Our main goal is to improve the service creation process for the homecare domain, in terms of reducing the IT skills, time, and effort needed to create new services, while aligning the individual needs of service users and the functionality of the created services. We call our approach to service creation "service tailoring". Service tailoring, as proposed in this thesis, is a way of creating new services, and adapting previously created services, involving healthcare professionals (care-givers) in the creation process and targeting elderly people (care-receivers) as the primary users of the created services. For requirements engineering, we started with a literature study of existing homecare systems and user-centric service personalization techniques, and a market survey of current homecare technologies and products. Then, we performed five series of interviews with caregivers in a care institution in the Netherlands (http://www.orbisconcern.nl/). This institution consists of residential blocks where elderly persons can live and receive care services round the clock from professional care-givers. The first, second, and third series of interviews were done before prototyping our approach, and the fourth and fifth series of interviews were performed to check the usability of the approach after the prototype was used by the care-givers and care-receivers. We performed the first series of interviews by using questionnaires to analyze the existing situation and to learn what types of services were expected from an IT-based homecare system. These interviews were helpful to gain insight into commonly performed tasks and how these tasks are performed. After designing our approach, we performed the second series of interview with the same care-givers to validate our design and refine it before implementing a prototype of our approach. Use of new IT-based services can introduce new types of risks in the environment where these services are implemented. This is particularly true in the homecare domain, since people’s wellbeing and health may depend on the services. Risks can possibly lead to unwanted or dangerous situations, and lack of trust in the services, e.g. due to multiple risk occurrences, may lead to a decline in the use of the IT-based services in practice. We proposed a Risk Driven Requirements Specification (RiDeRS) method to identify potential risks of using IT-based services, and to specify additional requirements on these services (and the underlying IT system) to mitigate or prevent these risks. To define RiDeRS, we performed a literature study of existing requirements elicitation methods which use risks to elicit risk-reducing requirements. Comparing RiDeRS with existing methods, what is new about RiDeRS is that it takes into account the properties of the system’s environment more systematically than existing methods. In RiDeRS, we considered users’ properties in addition to their goals to identify a list of possible risks and specified the requirements which can prevent or mitigate the risks. Accordingly, we performed risk assessment using RiDeRS by performing the third series of interviews with the same care-givers to identify and analyze potential risks of using our services and consequently to identify further requirements of the services and the underlying IT system. After evaluating the result of the first three series of interviews, we designed the architecture of a service tailoring platform. To evaluate this architecture, we developed a prototype of the service tailoring platform as part of the U-Care (http://www.utwente.nl/ewi/ucare/) project, which was subsequently used in two series of experiments to validate the properties of the approach. The experiments were conducted in a near real-life setting at the care institution. A number of users (8 care-receivers and 4 care-givers) volunteered to use the U-Care system. We assumed that service tailoring should be done by a care-giver. We identified different types of care-givers who interact with and help care-receivers in their daily life in the homecare domain. The identified care-givers include professional nurses, family members, informal caregivers (volunteer non-professional care-givers), occupational therapists, physiotherapists, physicians, pharmacists and psychologists. We found that a professional nurse, as a care-giver, is the one who should tailor the services, as care-receivers spend more time with professional nurses than other types of care-givers. We aimed to achieve IT-transparency by using the concept of service building blocks (SBBs). The SBB concept is used to denote the smallest manageable unit of service functionality from the point of view of care-givers (e.g., a reminder SBB notifies care-receivers to do something). A SBB provides a generic service interface to be used by care-givers in the process of service tailoring. A SBB also provides a list of configuration parameters to allow a care-giver to specify different aspects of the SBB, such as service operations and user interface modalities. Each SBB represents a concrete service or alternative concrete services, abstracting away the technical details that are not relevant to service tailoring (i.e., irrelevant to the care-giver). The outcome of a service tailoring process is a so-called service plan, which represents a composite service tailored to the needs of a specific care-receiver as understood by the care-giver. A service plan contains sufficient information to allow the automated derivation of a complete implementation on a target execution platform. Designing such a service plan from scratch is a difficult and a time consuming task. We use the concept of a design pattern to simplify the process of creating a service plan. We make use of treatment patterns as a starting point for the tailoring process, where a treatment pattern is an activity structure for handling a generic homecare task (e.g., blood pressure monitoring task). Thus, the care-giver does not have to create a service plan from scratch, but rather selects the homecare task to be supported from a menu. The tailoring platform then presents the corresponding treatment pattern as the initial service plan, which should be further refined and completed by the care-giver. The service tailoring platform is responsible for enhancing the creation and tailoring of the service plans by providing a graphical user interface (GUI) to the care-givers. To show the feasibility of the proposed architecture, we developed a prototype of the tailoring platform. The prototype was evaluated in two series of experimental field tests (with a total duration of 4 months). After the first series of experiments, we performed a fourth series of interviews with the care-givers, evaluated the results and improved the system. After the second series of the experiments during which the users used the improved system, we performed the last series of interviews with the users to evaluate the usability and usefulness of the service tailoring platform. We also asked their opinion of the system in order to see if such a system could be used in practice and indeed if it could save care-givers time and could increase the quality of life of the care-receivers. Finally, we reflect on the work done and the results achieved in the context of the homecare domain, and subsequently discuss whether the proposed approach can be generalized for use in other homecare applications.