Overview of CARE
CARE is an assistive healthcare platform developed by the Center for Ubiquitous Computing at the University of Oulu. The platform consists of a innovative sensor infrastructure that collects contextual data from nursing environments using the AWARE framework. By applying state of the art machine learning algorithms, the data is converted into understandable information for nurses to use in the daily care service.
The project is divided into four distinct work packages with iterative design methodology:
- In WP1 we took residence at a local elderly care center, to understand the needs of both the elderly and nurses. Based on this study we formed a set of elderly and prototype requirements for our solution.
- For WP2, we started developing our sensor infrastructure, cloud service and Android application, while continuously evaluating our work with the nurses. The aim was to develop a tool which could provide value to the nurses and easily integrate into their workflow.
- In WP3, the CARE API was introduced. It offers an easy to use interface to the healthcare data collected by the sensor system, allowing third parties to access relevant data and embed it into their own services.
- WP4 involves integrating our solution into a wider range of services. A integral part of this is the commercialization of the system in collaboration with Haltian Oy.
Globally, society is ageing at a fast pace, with the population aged 60 and over, representing the fastest-growing group. In Finland, 27% percent of citizens are over 60, and this number is expected to grow to more than 32% by 2050. Due to different kinds of losses (e.g., death of loved ones, impairing illness), loneliness/social isolation becomes a severe threat, leading to major health problems among the elderly. Deficiencies in social relationships increases risk of developing coronary heart disease (CHD) and stroke, and loneliness is comparable with well-established risk factors for mortality such as smoking and alcohol consumption. Requiring constant attention, between 65-74 year olds, 20,3% of Finns report severe limitations in performing daily activities, higher (29,7%) for 75-84 year olds. This affects family’s stability, increases caregivers’ workload and leads to a severe upshot on the country’s health expenditure. According to an OECD report published in 2015, the Finnish health system is chronically underfunded, spending approximately €3,000 per year, per capita, on health care services. To fight elderly loneliness, since the beginning of 2016, students are given cheaper rents in Finnish seniors home around Helsinki, in exchange for participation on a variety of activities with cohabiting seniors, such as playing music or cooking. This initiative improved drastically elders’ mood and allows them to nurture new social bonds.
CARE’s objective is to take advantage of smartphones’ popularity and wearables’ sensing capabilities to fight loneliness/social isolation, help reduce caregivers’ workload and thus elderly care costs, by means of context-awareness. Specifically, by capturing fine- grained data of non-critical, non-intrusive health-related physiological measurements, e.g., skin temperature to detect fever and sleeping quantity; accelerometer to detect physical activity and sleep quality; skin conductance to detect pain and stress; and Bluetooth Low-Energy to detect the proximity of others and social interaction, we collect invaluable data on different wellbeing and social conditions, useful to enable context-aware elderly-oriented applications and services. As we inform caregivers and family members of wellbeing and social measures (e.g., insufficient social interaction, feeling feverish or stressed), we may encourage and inform social engagement needs with the elders, improving elders’ quality of life.
CARE Sensor Infrastructure
At the lowest level, the system consists of a sensor infrastructure, responsible for measuring patient and nurses activity levels, location and social proximity. It also measures various physical events in the care facility, such as bed movement, room temperature, humidity changes due to showering and toilet visits. This data is then collected by automation control devices, which uploads the data to the cloud for storage and processing (i.e., pre-processing, statistical summaries, machine learning). The analysis is then synced to our Android application, to provide the nurses with easy and understandable information. We deployed and evaluated CARE in an elderly care home.
All the collected sensor information is visualised and presented to the care workers through our Android application. This can run on either smart phones or tablets, which enables nurses to have access to this information on a mobile format. The main view displays a list of the patients, and their most urgent care needs. In combination with sensors data, we also allow the nurses to input their own information about the patients (e.g., medical, family, personality). This allows new worker to quickly learn about the patients and their status. In the health history tab, the nurses can see health changes over long periods, revealing interesting trends that can be hard to notice on a day to day basis. The application also has functionality to support nurses during the daily and highly critical handover meeting.
The CARE metrics shown in our application is carefully selected after validation studies in collaboration with nurses working in elderly care facilities. We know exactly what information nurses need in their work to deliver the best possible care service. When a patient has slept poorly, not been to the toilet in a while or have not been social a specific day, week or month, or application will let the nurses know.
- Simon Klakegg, Chu Luo, Jorge Goncalves, Simo Hosio, Vassilis Kostakos (2016): "Instrumenting smartphones with portable NIRS", ACM UbiComp (https://dl.acm.org/doi/10.1145/2968219.2971590)
- Simon Klakegg, Jorge Goncalves, Niels van Berkel, Chu Luo, Simo Hosio, Vassilis Kostakos (2017): "Towards Commoditised Near Infrared Spectroscopy", ACM DIS (https://dl.acm.org/doi/10.1145/3064663.3064738)
- Simon Klakegg, Niels van Berkel, Aku Visuri, Chu Luo, Jorge Goncalves, Simo Hosio, Hanna-Leena Huttunen, Denzil Ferreira (2017): "Informing caregivers through an assistive tool: an investigation of elderly care metrics", ACM British HCI (https://dl.acm.org/doi/10.14236/ewic/HCI2017.46)
- Simon Klakegg, Niels van Berkel, Aku Visuri, Hanna-Leena Huttunen, Simo Hosio, Chu Luo, Jorge Goncalves, Denzil Ferreira (2017): "Designing a context-aware assistive infrastructure for elderly care", ACM IMWUT (https://dl.acm.org/doi/10.1145/3123024.3124403)
- Hanna-Leena Huttunen, Simon Klakegg, Nils Van Berkel, Aku Visuri, Denzil Ferreira, Raija Halonen (2017): "Understanding elderly care: a field-study for designing future homes", ACM IIWAS (https://dl.acm.org/doi/10.1145/3151759.3151835)
- Simon Klakegg, Jorge Goncalves, Chu Luo, Aku Visuri, Alexey Popov, Niels van Berkel, Zhanna Sarsenbayeva, Vassilis Kostakos, Simo Hosio, Scott Savage, Alexander Bykov, Igor Meglinski, Denzil Ferreira (2018): "Assisted Medication Management in Elderly Care Using Miniaturised Near-Infrared Spectroscopy", ACM IMWUT (https://dl.acm.org/doi/10.1145/3214272) *Distinguished Paper Award*