The objective of the Smart Home and Assisted Living (SHAL) Demonstrator is to illustrate the benefits of a big data platform through the implementation and offering of a services bundle towards advanced holistic monitoring and assisted living management, which aims to aid the everyday wellbeing of people belonging to vulnerable groups. In summary, the case study is the following: a social care service provider, for example a care centre for elderly individuals or a nursing home, desires to exploit big data-driven insights, in order to provide added value services to vulnerable individuals. The services pertain proactive and reactive security and protection through smart notifications and personalised recommendations, as well as indoor comfort and quality preservation. Proactivity and reactivity of the aspired services aim at prolonging self-sufficiency and independence of the at-risk individuals, boosting safety, and facilitating informed decision making, either by the individuals themselves, or by their (in)formal carers. The demonstrator is developed in Athens by Hypertech, UBITECH and Suite5, all Information and Technology (IT) companies which adopt the AEGIS roles of the service developers and data scientists.
Overall, the SHAL demonstrator will implement two main services that can be offered by a care service provider to at-risk individuals and/or their (in)formal carers:
- Monitoring and analysis of an individual’s well-being conditions, physical activity, positioning and wearable information and external environment data (e.g. weather, crime, news, social media), towards provision of a service for personalised notification and recommendation system for at-risk individuals, including notifications for carers.
- Additional service pertaining monitoring and analysis of weather, indoor environmental conditions, energy and operational device data towards the provision of a smart home application, which can be offered by care providers to at-risk people for increased indoor comfort and welfare.
These services were broken down to six implementation scenarios, three for each service pertaining the early, medium and advanced version of the demonstrator, following the time plan for the circulation of versions 1, 2 and 3 of the AEGIS platform. Here we discuss the results from the implementation and evaluation of the early demonstrator scenarios:
Scenario 1 – (At-risk) Individuals data fetching, processing and classification
Scenario 1 includes the steps that have to be performed in order to onboard an individual and a carer to the system, which results in the acquisition of personal data that are used to classify (at-risk) individuals into specific profiles (personas). The accumulated data is then used (in an anonymous manner) by the Care Service Providers (CSPs), in conjunction with external data, for further analysis and for issuing simple notifications. In parallel, cares are registered to the platform and are connected to the (at-risk) individuals.
Seven test cases were implemented to evaluate the functionality prescribed in Scenario 1. Specifically, the necessary registration mechanisms were implemented in the SHAL Web App, with a simple profile page created where each user (either an at-risk individual or a carer) would be able to specify some personal information, as well as some information regarding his health and his conditions, and register devices to the platform, so as to allow the system to perform simple queries and categorisation functions. Furthermore, the application allows individuals to view available care takers and connect their profiles. On the Aegis platform, the Algorithm Execution Container was utilized to create a custom cluster analysis of at-risk individuals into specific profiles (Personas). The profile data of the at-risk individuals were first anonymised by removing any sensitive data, thus ensuring that the individual records or subjects of the data cannot be re-identified. In addition, the Aegis event detection toolkit was utilized. Events that are of interest for the CSPs and relate to their patients have to be identified so that the CSP Analyst can decide on issuing notifications to his monitored individuals. A notification mechanism reaching both individuals and also care-takers (not included in the original test case) has been implemented, using FireBase which allows the real-time provision of notification to users of mobile devices. Finally, notifications are issued by CSP analysts and are sent to person groups, where the CSP analyst only knows the user-id of people belonging to a persona, and not their real personal data. Once a notification is sent, a copy of it is sent also to the care taker who is linked to an individual.
Scenario 2 – Smart home data monitoring and processing
Scenario 2 constitutes the early version demonstrator with relation to the added value service of smart home automation. The scenario includes the following user stories and associated test cases, from the perspectives of the CSP and at-risk individuals respectively.
CSP: The first step towards the implementation of the smart home offering, pertains the establishment, by the data scientist working for the CSP, of data flows regarding the at-risk individual’s indoor conditions, and associated external weather measurements, as well as the required pre-processing and normalization that will allow, in a subsequent step, to train the profiling mechanism and estimate the personal preferences of the individuals.
At-risk individual: The individual, after registration to the SHAL service, is supplied with a mobile application. The Smart Home monitoring UI allows real-time information on temperature, humidity, VOC concentration and HVAC status to be visualised for informative reasons.
Another three test cases were defined for evaluation of scenario 2. The first pertained the establishment of the necessary hardware and software infrastructure for the collection of the defined smart home data streams. To achieve this target, a number of sensing and actuation instruments were installed. In particular, a proprietary multi-sensor and gateway combo device, developed and manufactured by Hypertech was installed. The combo device is equipped with luminance, temperature, humidity and VOC sensors, as well as the gateway controller, which accumulates the signals and transmits them to a server through a rest interface, using Wi-Fi and TCP/IP communication protocols. The combo device was coupled two other actuation instruments.an air condition actuator module, which allows monitoring and control of the HVAC device and a dimmable ballast with wireless controller for measuring and setting the dimming level on lighting devices. The monitored data are sent directly to the SHAL backbone server through a secure Rest interface. No personal data are transmitted, apart from the asset id, whose details were provided by the user upon successful registration to the demonstrator’s app. The second test case concentrated on the evaluation of the data processing capabilities of the AEGIS platform. Datetime entries were transformed to a standard format, and all data points were sorted according to their measurement time. Missing values for all measurement variables were filled in. Interpolation of all measurements to a constant time frame were also performed, and finally, motion data from the PIR sensors were processed in order to extract binary occupancy data. For the final case, a preliminary subset of the visualisations of smart home data, for informative purposes were implemented in the UI app.
All test cases were successfully completed. Implemenation and evaluation continued without significant issues towards the delivery of hte second version of the demonstrator. More details can be found in the corresponding deliverable https://www.aegis-bigdata.eu/wp-content/uploads/2017/03/AEGIS-D5.3-Demonstrators-Evaluation-and-Feedback-v1_v1.0.pdf.