The context of this project is the growing need for automated health-care support for the sick and the elderly at home that can provide them with more comfort and safety while reducing the overall cost of the health-care system. Smart spaces have the potential to provide such automated support as numerous projects have demonstrated [Anliker04, Moulton04, Quarrie07, Baker07, Gonçalves09] and as a recent European research funding initiative on Ambient Assisted Living emphasizes (http://www.aal-europe.eu/).
The problem of interference
Smart spaces (also known as pervasive or ubiquitous computing environments) can be seen as a collection of disperse sensor, actuator, computation, and communication-rich appliances and infrastructure in which new functionality can be added by bringing in new appliances. An important aspect in managing smart spaces is to understand the extent to which a new appliance that is brought into a smart space can interfere with the appliances that already exist in that space [Edwards01, Davies02, Morla06]. This is known as the feature interaction problem in the telecom domain [Calder03], where features that are added to the telephone network to make it smarter (e.g. call forwarding) can interact with each other in unexpected ways. Addressing interference in smart spaces for home health-care is especially critical because of the human life factor; e.g., an appliance may be prevented from transmitting important health data because of electromagnetic interference generated by other appliances; an appliance body temperature-monitoring may be affected by heating appliances; and the users may be distracted from taking their drugs by a multimedia appliance.
Although the problem of interference is critical for smart spaces in general and home health-care spaces in particular, controlling interference in these spaces is not obvious [Edwards01, Davies02]. Different solutions exist for addressing specific forms of interference – the best known of which is probably electromagnetic interference. In smart spaces however, interference must be considered across different layers – from physical to networking, application, and user – and between highly dynamic appliances, which makes the problem considerably more complex. Interference has been studied before specifically for smart-spaces [Gajos01, Kolberg03, Shankar06, Morla07a, Tuttlies07], yet without yielding fully satisfactory results. This is mostly because of the focus on specific layers of the smart space [Gajos01, Kolberg03, Shankar06], of considering resource access and not appliance behavior [Gajos01, Kolberg03, Tuttlies07], and of the focus on e.g. user study design and not on smart-space development [Morla07a].
A novel approach
The goal of this project is to develop and evaluate a novel approach based on meta-reality reflection and search algorithms that helps finding and deploying interference-free health-care smart-space configurations. To achieve this goal, we will:
1) use reflection to dynamically reify the structure and behavior of smart spaces into a virtual-reality meta-model and reflect changes in the meta-model back to the smart-space;
2) use search algorithms to efficiently navigate the meta-model configuration space and find interference-free configurations of health-care smart-systems.
This project will have three tasks. The first two tasks will have a conceptual contribution by analyzing and designing the dynamic reification-based meta-reality parametric model and the use of search algorithms for finding interfere-free configurations. Tools based on virtual world software will also be produced to support these contributions. Although simple use cases defined in these tasks will help in the design and evaluation, the validation of these contributions will mostly occur in task 3 in which a real health-care smart space will be designed and deployed.
The project team has experience in the cross-cutting issues of smart-space deployment and interference [Morla06, Morla07a], in virtual worlds [Morla07b,Lopes08], in reflection [Moreira04], and health-care [Gonçalves09]. Recent projects of the PI include a FLAD-NSF-funded project on prototyping pervasive computing using virtual worlds with UC Irvine and an ADI-funded project in the Rede de Competência em Mobilidade on interfacing wearable health-monitoring devices with virtual worlds.
The research results of this project are to find out to which extent this approach can address interference and is better than existing approaches, and how it performs in a home health-care smart space. As the average age in Portugal and the World becomes higher, many Portuguese companies have been investing in home health-care and entertainment for the home (e.g., BioDevices, YDreams) and we believe the scientific results of this project could be used for the benefit of the expansion of these companies’ national and international markets.