The intelligent habitat and everyday life activity support Hlne - - PowerPoint PPT Presentation

1 de 32

The intelligent habitat and everyday life activity support

Hélène Pigot, André Mayers, Sylvain Giroux Département de mathématiques et d’informatique, Université de Sherbrooke, Canada. {pigot, andre.mayers, sgiroux}@dmi.usherb.ca

2 de 32

Introduction

The need for intelligent habitats in upholding elders in residence

Their number increases They wish to remain at home Governments want to maintain them in their residence. Elders are suffering from several chronic diseases

The safety problems

Immediate risks Long-term risks

With the current technology in perspective, we show how the habitat becomes active and intelligent

to assist elders in their Activities of Daily Living (ADL) to inform relatives and caregivers as soon as necessary.

3 de 32

Outline

Introduction (done) Description of the target population Overview of the proposed solution Computer infrastructure Implementation Cognitive modeling aspects Scenario Conclusion

4 de 32

The Global Deterioration Scale for Assessment

• f Primary Degenerative Dementia

Continuous care and monitoring are compulsory to keep at home elders suffering from dementia This burden is too often simply delegated to their relatives. As a result relatives frequently suffer from distress, fatigue…

1 No cognitive decline no subjective or objective deficits 2 Very mild cognitive decline some sujective complaints, no objective deficits 3 Mild cognitive decline mild working memory deficits (attention, concentration) 4 Moderate cognitive decline episodic memory deficits (memory of recent events) 5 Moderately severe cognitive decline explicit memory deficits (ability to accomplish usual tasks) 6 Severe cognitive decline severe memory deficits (which cause delusion) 7 Very severe cognitive decline all verbal activities are lost

5 de 32

Outline

Introduction (done) Description of the target population Overview of the solution proposed Computer infrastructure Implementation Cognitive modeling aspects Scenario Conclusion

6 de 32

Physical environmental strategies

To inform family or medical staff of disease evolution, and critical situations To provide appropriate environmental cues* in order to help elder in their activities of daily living (ADL) to diminish risks at home

*according to minimal intervention principle

As people must use their capacity in order to maintain it therefore If the elder can do it safely without cues the system won’t intervene

7 de 32

Recent progress in science and technology

pervasive computing ubiquitous computing wearable computing wireless networks mobile code intelligent sensor

will drastically transform houses Now we propose a pervasive information system

to gather information from the environment to analyse it to intervene according to people needs and preferences Approaches distance monitoring cognitive assistance. and then distributed artificial intelligence agent

8 de 32

Symbiosis between monitoring and cognitive assistance approches

action intelligent appliance sensor data analysis processes psychological, medical, environmental model report disease evolution alert external support to enhance security and confidence generate appropriate to support or enhance the individual’s performance

Telemonitoring Cognitive assistance

environmental cues goal goal action use almost the same

9 de 32

Ethics

We don’t pretend and don’t want to replace human communication We want to alleviate distress and fatigue for relatives. We want security, autonomy and human dignity for elders.

10 de 32

Adapted interventions

Characterized

domain: hygiene, alimentation … severity: “doing nothing”, advice, drastic frequency: low …

Personnalized

cognitive capacities: decreasing rate of memory trace activation cognitive abilities: “cooks spaghetti sauce wihout any hesitation” habits: nap after lunchtime … preferences: acoustic, visual signal … history: “eat chicken for lunch today”

11 de 32

Outline

Introduction (done) Description of the target population Overview of the solution proposed Computer infrastructure Implementation Cognitive modeling aspects Scenario Conclusion

12 de 32

A layered computer infrastructure

computer micro-wave bathroom door kitchen table glycometer hardware layer code middle layer code Interpretation and decision making layer code integrator cognitive assistance module telemonitoring module module

13 de 32

Outline

Introduction (done) Description of the target population Overview of the solution proposed Computer infrastructure Implementation Cognitive modeling aspects Scenario Conclusion

14 de 32

Implementation

The support system will open out in very diverse contexts ad hoc specification of decisions and actions is neither practical nor feasible the system must provide models to describe: what to do, how to do it, and when to do it. The first prototype The Kitchen Task Assessment

• ne of the canonical ADL in occupational therapy

complex cognitive challenges home risks

15 de 32

Outline

Introduction (done) Description of the target population Overview of the solution proposed Computer infrastructure Implementation Cognitive modeling aspects Scenario Conclusion

16 de 32

The ADLs support system

Person meta-model Cognitive physiological behavioral Activities of daily living (ADL) scripts Activity meta-model Environment meta-model Sensors appliances visual, accoustic interfaces Instantiated person Activities in process Instantiated environment Decision and action: competence or handicap integrator module

Intelligent habitat

with a real person cognitive assistance module distance monitoring module middle layer code Competence model (Rousseau)

17 de 32

Outline

Introduction (done) Description of the target population Overview of the solution proposed Computer infrastructure Implementation Cognitive modeling aspects Scenario Conclusion

18 de 32

• Mrs. Smith’s afternoon scenario

Autumn has come, afternoons are colder and darkness falls sooner. Mrs. Smith likes to take coffee at 4:00 p.m. On Friday afternoon, as every week, Mrs. Smith decides to cook a veal

• stew. Water is boiling for coffee and the veal

stew is simmering when the phone rings.

HIT

• Mrs. Smith’s

19 de 32

• Mrs. Smith’s afternoon scenario

Inte gr A s s i s t a n c e T e l e m

• n

i t

• r

i n g midd ware le a t

• r

Computational model of

• Mrs. Smith’s

Computational model of

• Mrs. Smith’s

cognitive and physiological aspects activities Computational model of

• Mrs. Smith’s

HIT HIT

• Mrs. Smith’s

as the system infers them

20 de 32

21 de 32

Identification Female 73 years old

Alzheimer disease and osteoporosis

Cognitive abilities SAS/CS: unable to coordinate more than 2 middle tasks working memory: severely impaired, unable to

remenber more than 3 items for 30 seconds.

episodic memory: mild, remember many facts from

• ne day before

explicit memory: no deficits Habits

Light meal taken at 4:00 p.m. Dinner at 6 p.m.: soup, meal, desert. She likes to cook boiled meat, twice a week

Preferences

Acoustic signal

Computational model of

• Mrs. Smith’s

cognitive and physiological aspects

22 de 32

Physical

• 1. equipment available
• 2. current state

Stove equipped with graphic and acoustic interface with

turn off process automated and selective Sensors in the kitchen, in the lounge and the bedroom Sensors on: 2 stove units, detection of presence Sensors off: bedroom, hall, lounge Telephone: ringing Human Daughter to call during the evening at home

Doesn't like to be disturbed too often

Medical staff to inform every month about the evolution

and in case of emergency

Computational model of

• Mrs. Smith’s

HIT

=/=

HIT

• Mrs. Smith’s

23 de 32

Cooking activity 1 Certainly, she is preparing a meat stew

evidence: sensors, habits, time of day (script1 0,45 ; script2 0,35 ; script3 0,20)

Cooking Activity 2 Certainly, she is preparing a hot coffee

evidence: sensors, habits, time of day (script6 0,75 ; script7 0,25)

Telephone activity Certainly, she will respond to the phone call

evidence: ringing, habits (script17 1,00)

Computational model of

• Mrs. Smith’s

activities as the system infers them

24 de 32

Inte gr A s s i s t a n c e T e l e m

• n

i t

• r

i n g midd ware le a t

• r

HIT

• Mrs. Smith

‘ s

Warning 3 activities are in process Difficulties to cope with more than two activities Cooking activities are fire risk

25 de 32

• Mrs. Smith’s afternoon

Ten minutes later

Ten minutes later, Mrs. Smith has finished her phone call. As she feels tired, she decides to rest for a while in her bedroom.

HIT

• Mrs. Smith’s

26 de 32

Physical Sensors on: 2 stove units , bedroom Telephone: off, telephone is put down Sensors off: kitchen, hall, lounge Human Daughter

To call during the evening at home Doesn't like to be disturbed too often

Medical staff

To inform every month about the evolution and in case of emergency

Computational model of

• Mrs. Smith’s

HIT

27 de 32

Cooking activity 1 Certainly, she is preparing a meat stew

evidence: sensors, habits, time of day (script1 0,45 ; script2 0,35 ; script3 0,20)

Cooking activity 2 Certainly, she is preparing a hot coffee

evidence: sensors, habits, time of day (script6 0,75 ; script7 0,25)

Phone activity Certainly, done,

evidence: lounge sensor off, …

Nap Maybe, she is resting or reading journal

Computational model of

• Mrs. Smith’s

activities as the system infers them

28 de 32

Inte gr A s s i s t a n c e T e l e m

• n

i t

• r

i n g midd ware le a t

• r

HIT

• Mrs. Smith

‘ s

Suddenly, the system fire alarm triggers

Immediatly, the stove turns off all its components Two informations are sent to the upper level

• fire alarm activated
• stove turned off

29 de 32

Inte gr A s s i s t a n c e T e l e m

• n

i t

• r

i n g midd ware le a t

• r

HIT

• Mrs. Smith

‘ s

Warning 3 activities are in process Difficulties to cope with more than two activities Cooking activities are at risk of fire

30 de 32

Environment Person Activity Decision Mild risk

• f fire

Mild deficits Activities still in control 1 .Warn the person to turn off Moderate deficits Just boiled water dangerous

• 2. Put the veal stew on

Severe risk

• f fire

Moderately severe deficits Activities too dangerous

• 3. Let the stove off

Inte gr A s s i s t a n c e T e l e m

• n

i t

• r

i n g midd ware le a t

• r

HIT

• Mrs. Smith

‘ s

31 de 32

Outline

Introduction (done) Description of the target population Overview of the solution proposed Computer infrastructure Implementation Cognitive modeling aspects Scenario Conclusion

32 de 32

Conclusion

Two kinds of interventions Model of the relationship between the person, the activities and the environment. Distributed architecture One ADL is under study, the meal preparation activity The scenario