The Future of Software Development - Devin AI Innovative Approach.pdf
4 subasi ehome-ifa-prague-v03 short
1. Experiences from Assessing Daily
Activities to Increase Safety and
Comfort of Older Persons Living Alone
Paul Panek, Peter Mayer, Özge Subasi
and Wolfgang L. Zagler
Institute for Design and Assessment of Technology
Vienna University of Technology
IFA 11th Global Conference on Ageing
28 May – 1 June 2012, Prague, Czech Republic
3. Introduction & Aim
Many research papers available about AAL
systems and activity monitoring
– But unclear, to what extent such systems can be
applied in real life
Therefore - Aim of eHome project:
– Demonstration of practical usefulness
– In daily use
– Basic system with ZigBee based sensors was
developed
4. Approach
Daily life follows certain schedules
– Given by routine
– Following social and biological rhythms
eHome monitors continuously
– Assessing activities in fixed time slices (e.g. 1 hour)
– Triggering supportive or emergency actions if
significant deviations
– Keeps private data protected inside the user’s home
Objective of eHome: Improving safety and
comfort of older persons living alone at home
5. Approach
eHome system uses
– Sensors connected via ZigBee
– Storing data in a database
– Data being processed in a small central
unit
– Situated at user’s home ( privacy of
data)
– Connected to the Internet.
6. Approach
Event triggered expert system can raise
alerts based on:
– Time of activities
(e.g. rising from bed compared to daily history)
– Duration of selected activities
(e.g. nightly leaving of bed)
– Frequency of activities
(e.g. reduction in cooking, hygiene)
7. Approach
Assumption of eHome is:
– Even by applying a rather coarse monitoring by
a small set of sensors a sufficient insight into the
user’s activity can be reached
– Even if “better” and “more” sensors would be
possible this was avoided in order to
Increase perspective for economic exploitation
Improve the to-be-expected user acceptance
8. Field Trials
Extensive evaluation with
– 5 research prototype systems
– installed in 11 homes of older persons
– over a total time of 18 months
9. Field Trials
ZigBee Sensors for
– Door / window
– Acceleration & floor
vibration
(e.g. for fall detection)
– Temperature
– Cooking plate temp.
(infrared)
– Movement (passive
infrared)
– Light
10. Impressions from Field Trials
Local User Interface
– Touch-Screen Terminal
without typical PC look
– Fits to furniture
– Easy to use
– Video telephone
– Smart-Home Controls
– Emergency Call
– Internet Browser
14. Results
System was able to classify “usual
behaviour” over time. This can be used in
different ways:
– Unusual sudden changes (e.g. not leaving bed in
the morning) triggering alarm
– mid-term and long-term trends present
changes to care persons to let them judge about
the meaning and severity of recognised
changes.
16. Results
The system is able to learn / adapt over time its
threshold parameters
Remark: Even right from the beginning the system
is able to work with initial values important for
practical use!
By adapting over time it will improve performance
17. Conclusion
Despite needs for improvement there is evidence
that the system actually is considered by users and
experts to have the potential to bring significant
benefits in supporting older persons and carers.
eHome prototype currently is used in LLM (CIP)
and KSERA (FP7) project
Costs for low quantities of basic eHome system
with 3 multi-sensor boxes are 1,500 Eur.
19. Acknowledgements
Supported by: Austrian Federal Ministry for
Transport, Innovation and Technology
(FIT-IT contract number 815195).
Consortium: TU Vienna, Ceit Raltec, Kapsch
Carriercom and Treventus Mechatronics.
For questions: subasi@igw.tuwien.ac.at
Web site: www.aat.tuwien.ac.at/ehome/