Towards Medication Management in Smart Homes

Andrej Grgurić

, Saša Dešić

, Miran Mošmondor

and Mario Kušek

Ericsson Nikola Tesla, Krapinska 45, 10000 Zagreb, Croatia

Faculty of Electrical Engineering and Computing, University of Zagreb, Unska 3,10000 Zagreb, Croatia

Keywords: Medication Management, Smart Home.

Abstract: Due to current demographic trends and ageing more and more people are living alone and need proper

support in their daily activities. We propose a model for smart home environment upon which we base our

solution described in this paper. Example solution aims to increase reassurance for people living alone by

providing them with the reminding and smart home functionalities encapsulated in terms of OSGi bundles.

Medication management and timely notifications combined with lighting and audio system control allow

elderly users to individually increase their autonomies while living alone.

1 INTRODUCTION

According to Eurostat there is a trend of

demographic ageing within the EU. The share of

older people is drastically increasing from 2010

onwards because of a low fertility rate and rising life

expectancy. It is estimated that the ratio of working-

age people to those aged over 65 will drop from 4:1

in 2008 to less than 2:1 by 2060.

With the advancement of the technology and

higher penetration of broadband networking a vast

number of new possibilities in the area of intelligent

environments have emerged. On the one hand there

are new ways for collecting and storing data and on

the other hand there are new means for interpreting

and reasoning upon that data which, in addition to

the improvements in human-computer interaction,

greatly facilitates the advancement and development

of different inventive systems. Those systems further

enable development of context-aware applications

with new ways of adaptation, personalization and

customization that has never before been possible.

Users can now choose among ever increasing

number of different solutions which can ease their

everyday tasks and provide them with support they

need. However, current solutions still have many

limitations and thus are not well established.

Usage of proprietary, legacy and isolated

technologies hinders the advancement directed

towards reusability and interoperability and hence

also towards cost effectiveness and widespread

acceptance. Nevertheless, in the last years a lot has

been done in that sense, which can be seen also on

the example of initiatives such as Continua Health

Alliance, an open non-profit organization of more

than 230 technology and healthcare companies, or

Digital Living Network Alliance (DLNA),

organization focused on delivering an

interoperability framework of design guidelines

based on open industry.

Today there are many ongoing research efforts

directed towards building intelligent home

environments such as Gator Tech smart house

(Gator Tech, 2012) or Aware Home (Aware Home,

2012). Projects like MPOWER (Mpower, 2006) and

PERSONA (Persona, 2007) aimed at developing

technological platforms to enhance independent

lives of elderly. Mentioned projects are also input

projects into the project universAAL (UNIVERsal

open platform and reference Specification for

Ambient Assisted Living) (universAAL, 2010) which

has an objective to integrate the current state-of-the-

art projects and solutions into new Ambient Assisted

Living (AAL) reference platform.

The above mentioned projects and initiatives,

together with many similar ones, aim at utilizing

different devices and systems in a way that they

reduce the need of manual work of the elderly and

provide them with assistance in their daily lives.

Solutions provided by those projects are all about

employing technology in order to offer elderly

people new means of help and support.

Numerous reports show that patient

noncompliance, also known as non-adherence, is a

228

Grguric A., Deši

c S., Mošmondor M. and Kušek M..

Towards Medication Management in Smart Homes.

DOI: 10.5220/0004238002280232

In Proceedings of the International Conference on Health Informatics (HEALTHINF-2013), pages 228-232

ISBN: 978-989-8565-37-2

 2013 SCITEPRESS (Science and Technology Publications, Lda.)

major medical problem. Noncompliance is typically

cited as occurring in from 50% to 75% of patients

(Wertheimer, 2003). This rate is even higher in

patients with chronic illnesses since their drug

regimes are often long term. The economic and

healthcare costs are thus even higher not to mention

that a number of people die every year because of

not taking the proper drug dosages at the right times.

Automation is certainly one of the key aspects

that drive research in Smart Homes and, thus, also in

AAL domain. Smart environments can be examined

from the perspectives of work, leisure, safety,

comfort and privacy. By developing medication

reminder system coupled with smart home control

we wanted to a) demonstrate new ways of reminding

in smart homes, b) utilise different OSGi

implementations for developing applications and c)

transfer existing non-OSGi MPOWER FSA

framework to OSGi.

The rest of this paper is organised as follows:

first an example scenario and proposed smart home

model are described in Section II. Section III

continues with the elaboration of concrete system

architecture and Section IV concludes the paper.

2 SMART HOME APPROACH

2.1 Requirements

We will use a following example to illustrate how

even a pretty simple system can provide the assisted

person with a good reminding and smart home

functionality. Consider a 72 year old woman Agnes

who lives alone, suffers from dementia and has a

difficulty to move around. She has to take several

medications at appropriate, often different, times. A

few times she unintentionally forgot to take her

medications because of which she had had medical

problem and decided try to prevent this from

happening again.

She knew that a plethora of different services is

available and can be ordered online. In consultation

with her son she decided on the most suitable

solution for her problem which offers timely

reminders and some smart home functionality. She

bought it and after technician installed everything in

her house she asked her son Steve to help her by

using the part of the solution meant to be used by a

caregiver. Steve took over the task of entering

scheduled medication intake times to the system so

Agnes could continue with her daily activities more

carefree then before. By knowing that she will be

reminded to take medication on time she felt

reassured and relieved. Moreover, she knew that she

doesn’t have to be concerned whether she will see or

hear the notification on her mobile since the alarm is

integrated into her house and she could hear the

sound and/or see the lights blinking no matter which

room she is in at that specific moment.

As an addition to the reminding functionality she

also acquired a service for turning the audio system

and light on and off. This device helped her avoid

getting up all the time which made her very happy

because of the problems with her painful knees.

From this example scenario we can see that

feeling of reassurance is invaluable for the end user.

We are witnesses that nowadays, people are

generally more willing to invest into home

automation devices. Quite a few manufacturers are

already addressing this area so numerous solutions

for controlling and programming of heating,

lighting, doors, windows and other electrical

appliances already exist on the market. The next step

is to see how they can be properly utilised in terms

of providing appropriate support for elderly. Rapid

miniaturization and development of sensors,

actuators and computing devices greatly facilitate

the advance of such solutions.

2.2 Proposed Model

When transferring technology explored in laboratory

conditions into domestic spaces a lot of practical

problems emerge. A relatively simple thing as

sensor colours can greatly affect the acceptance of

pervasive technology in the home. Many examples

show that a trade-off between the amount of

technology that can be installed and the level of

support and usefulness of the system has to be made.

Designers of the novel services have to take under

careful consideration social and privacy issues as

well as technical ones.

Sensing and responding technologies should

harmoniously be incorporated in physical

environments. With continuous progress in sensing,

novel activity recognition methods are also being

investigated. Sensors placed in houses or worn by

the users are constantly becoming smaller and more

advanced which definitely leads to Weiser’s

prophetic vision of the future of disappearing

technologies (Weiser, 1991). Apart from collecting

the data from the environment the opposite direction

also has to be covered and this is where actuators

come in by causing changes in environments.

We present a model for smart home on

Figure

1 that takes under consideration physical,

logical and interaction aspects by addressing sensing

TowardsMedicationManagementinSmartHomes

229

and responding, smart home service functionalities

and interaction with the user.

Figure 1: Proposed smart home model.

Smart home services that offer support to the

inhabitants should maximise the utilization of

obtained data and infer as much as possible in order

to be able to provide both passive and active support

to intended user. However, not all services need to

be context-aware and collect data from the

environment nor they need to utilise actuators. The

amount of hardware or software used for particular

service depends solely on service developers and

their intentions, the extent of end user needs,

preferences, etc. Furthermore, the amount of

usefulness of the service cannot be easily measured

so it is up to end users themselves to decide if some

service is useful for them or not. Assistive

technologies can therefore come in many flavours

and can range from the simplest to the most

technically sophisticated. They can be designed in

countless ways considering (or not) things such as

social, technical, ethical, physical and physiological

factors. The most important thing is, however, that

they provide some means of assisting their end

users. Having in mind that people who benefit from

smart home services often have really specific set of

requirements services, and especially more advanced

ones, have to be personalisable in some way.

Collaboration with external service and

information sources (such as weather forecast) must

not be neglected and can, in some cases, prove to be

invaluable for the completeness of the service

offering. Reusability and different orchestration of

services however are not possible without the

common ground for interoperability so the

standardisation compliance aspects are especially

important in this segment.

Interaction with the end user has a pivotal role for

every system since it greatly affects the overall

impression and thus also acceptance of the system.

Often the interaction aspect is greatly

underestimated although there are many research

efforts and advances in the field of human-computer

interaction. With the advancement of the technology

more natural means of interaction (such as voice- or

gesture- based) become possible. Multimodal

interfaces offer many possibilities and much easier

interaction which is especially important when

dealing with the needs of elderly people who are,

more often than not, really not technically savvy.

2.3 Concrete Architecture

Proposed system architecture follows above model

which gives the basic understanding of the entities

of smart home environment and their relationships.

Developed system is built using the OSGi

technology which is an industry’s initiative for a

standardised open specification to deliver services in

different environments and to bring support for

modularity for software developed in Java.

OSGi platform consists of OSGi framework and

running components called bundles which are

basically JAR (Java archive) files with additional

manifest headers. OSGi facilitates componentization

and assures interoperability, two very important

aspects in the AAL domain.

Figure 2: System architecture.

Figure 2

shows basic system architecture. Interaction

of the components is illustrated in a way that it is

comparable to the proposed smart home model.

Medication management client (MM client) is

OSGi based Java Swing application used for setting

a reminder about certain medication for specific

assisted person. For this client a popular open source

OSGi platform developed by the Eclipse

organization called Equinox (Equinox, 2012) was

used.

HEALTHINF2013-InternationalConferenceonHealthInformatics

230

Apache CXF Distributed OSGi subproject

(Apache CXF, 2012), or cxf-dosgi in short, was used

for Medication management web service (MM WS)

component implementation. Cxf-dosgi implements

the Remote Services functionality using Web

Services. It enables service invocation in distributed

environment meaning that services running in other

JVMs can also be invoked. Feature we used for our

development was leveraging SOAP over HTTP and

exposing the Service over a WSDL contract. Using

this framework, one can develop and deploy web

services as OSGi bundles. Apache Tomcat was used

as a servlet container.

While MM client is mainly intended for formal

(meaning medical personnel) or informal (meaning

family members, friends, etc.) caregivers, mobile

based application is primarily meant to be used by

assisted person (AP). Mobile client application is

used for presenting medication notifications and

interpreting smart home controls. It is based on

Prosyst’s mBS Mobile OSGi platform for Android

(ProSyst, 2012) which is complemented by mBS

Mobile SDK which offers a number of Eclipse based

tools.

For interaction with actuators a Frame Sensor

Adapter (FSA) middleware from the project

MPOWER was used. FSA is used as a gateway

providing unified access to sensors and actuators

that use different data formats and different

communication channels. There are three levels of

abstraction in FSA:

 Adapters are mediators between real devices

and Virtual Sensors. They possess knowledge how

to send a message via specific communication

protocol. For each protocol there is a special FSA

Adapter.

 Virtual Sensors possess knowledge of how to

interpret specific messages coming from different

devices.

 Frame offers interfaces to associate Virtual

Sensors and appropriate Adapters.

For the purpose of this work needed FSA

components were transferred to OSGi and necessary

adjustments, such as implementation of specific

Virtual Sensor, were made in order to be able to use

it within our system.

Caregiver or a family member enters medication

name and time when specific assisted person needs

to take prescribed medication. This data then

becomes available for the assisted person via web

service.

Figure 3: MM client sends data to MM WS.

Figure 3 shows transfer of a list of patient and

medication related data between MM client and MM

WS. When the list is downloaded on the assisted

person's mobile phone it is immediately shown on

the screen while appropriate timers are set in the

background.

Figure 4: Sequence diagram with focus on Mobile client.

Figure 4 shows communication flow when mobile

client requests medication data from MM WS, sets

alarm(s) and shows notification to the user.

When the alarm is triggered user receives a

notification on a mobile phone screen. If the

confirmation is not received for some time a stored

action is triggered and performed until user confirms

the notification. In our case user (AP) gets a visual

and an audio stimulus announcing that the time for

medication intake has come. By acting on more than

one sense probability that assisted person indeed

obtains the timely reminder drastically increases.

User interfaces for assisted person were made as

simple and easy to use as possible, without

displaying too much unnecessary information,

having in mind requirements of this specific target

group.

TowardsMedicationManagementinSmartHomes

231

2.4 Need for a Reference Platform

Different platforms, systems and solutions try to

address many problems in providing support to

elderly people living alone. The question we are now

facing is how to reconcile these solutions to get the

maximum result. Current research contributions

prove that scalable and manageable OSGi

technology is very valuable for smart home

environments.

universAAL platform addresses many of the

issues important also for smart homes. It terms of

runtime support it adds goal based interoperability

between different devices and services (it uses three

buses, namely Service Bus, Context Bus and a User

Interaction Bus where each is responsible for

handling specific calls and events between

distributed nodes). In terms of support for

developers it offers Developer Depot, a place where

different tutorials, explanations and examples can be

found. And lastly it offers uStore which, in short,

can be compared with Apple’s App Store but here

used for requesting, offering and obtainment of AAL

Services. By building on top of OSGi universAAL

runtime platform also benefits from all OSGi

underlying functionality. Additionally it adds

automatic discovery and seamless communication of

the nodes running universAAL middleware.

By using one common platform within smart

homes a lot of issues in terms of interoperability,

reusability, standardisation, cost reduction,

simplification, reduction of time-to-market cycle,

etc. can be resolved much quicker.

3 CONCLUSIONS

There is no doubt that a major shift has occurred in

the way technology is used today. Many advances in

science and technology allow developing services

that were formerly not possible.

Since elderly people often tend to forget their

medication intake schedules, reminders that

incorporate the smart home environment were

developed to help to alleviate these problems.

Implemented solution follows proposed smart home

model and shows that the usage of open standards

and frameworks, such as OSGi, can greatly facilitate

development of cheaper, simpler and more open

AAL applications and services. By using the

environment, instead just fixed screens, to attract

user’s attention a better outcome and medication

compliance can be achieved.

It is obvious that there is a lot of room for fully

integrated and pervasive solutions in smart homes.

However further research is needed to fully

understand the needs of the elderly and, moreover,

how different technologies and solutions can help

them to live independently in their homes.

ACKNOWLEDGEMENTS

Used FSA framework was initially developed under

the FP6 project MPOWER (#034707) funded by the

European Union and was made open source under

the MIT license. This work was partly supported by

the FP7 project universAAL (#

247950).

REFERENCES

Gator Tech smart house, Gainesville, Florida,

http://www.icta.ufl.edu/gt.htm [Accessed: July 2012].

Aware Home. Georgia Institute of Technology,

http://www.cc.gatech.edu/fce/ahri [Accessed: July

2012].

MPOWER Project, Middleware platform for empowering

cognitive disabled and elderly, 6h Framework

Programme of the European Union, Grant Agreement

No. 034707, 2006-2009, www.mpower-project.eu.

Persona Project. Perceptive spaces promoting independent

aging, 6th Framework Programme of the European

Union, Grant Agreement No. 045459, 2007-2010,

http://www.aal-persona.org.

universAAL Project, UNIVERsal open platform and

reference Specification for Ambient Assisted Living,

7h Framework Programme of the European Union,

Grant Agreement No. 247950, 2010-2014,

http://www.universaal.org.

Equinox OSGi Platform, http://www.eclipse.org/equinox

[Accessed: July 2012].

Apache CXF Distributed OSGi,

http://cxf.apache.org/distributed-osgi.html.

ProSyst, mBS Mobile for Android,

http://www.prosyst.com/index.php/de/html/content/49/

mBS-Mobile-for-Android/ [Accessed: July 2012].

Wertheimer Albert I., Santella, Thomas M., 2003.

Medication Compliance Research: Still So Far to Go,

The Journal of Applied Research in Clinical and

Experimental Research.

Weiser, M., 1991. The Computer for the 21st Century,

Scientific American, 265(3):94–104.

HEALTHINF2013-InternationalConferenceonHealthInformatics

232