IoT-Enabled Community Care for Sustainable Ageing-in-Place - PowerPoint PPT Presentation

IoT-Enabled Community Care for Sustainable Ageing-in-Place Hwee-Pink TAN, Ph.D. Associate Professor of Information Systems (Practice) Academic Director, SMU-TCS iCity Lab 19 May 2017

About the SMU-TCS iCity Lab The iCity lab was established in Aug 2011 to explore and pursue new research areas in Smart Cities to provide long- term competitive advantage to TCS - i = {intelligent, integrated, inclusive, innovative} - Leverages TCS’s and SMU’s strength in IT and management In view of the successful partnership in the last 6 years, TCS has committed funding to extend the relationship for a further 3 years (Aug 2017 – July 2020) to take the iCity lab to greater heights 2

The iCity team Steering Committee PROF Steven MILLER Ananth KRISHNAN Girish RAMACHANDRAN V. Provost (Research), SMU CTO, TCS President & Head of TCS Asia Pacific Core Team Raghavan V Sanjoy Biswas ANG Chip Hong DON Wijay TAN Hwee Pink Head of iCity Business Development Senior Associate Director Assistant Manager Associate Professor of Information Systems Initiatives, TCS Asia Pacific (Practice) and Academic Director TCS CTO Office TAN Hwee Xian Nadee G Alvin VALERA Research Scientist Research Fellow Research Fellow Cheryl KOH TAN Lee Buay BAI Liming NG Boon Thai Pius LEE Community Coordinators / Research Assistants Research Engineer Research Engineer Senior Research Engineer 3

Elderly who live alone are at risk! 2x 1.7x Source: The Straits Times, 18 Dec 2015 More likely to More likely to feel depressed die prematurely 4

Technology Pilots & Services An initiative Response by from staff/ community ~23,000 elderly homes Source: The Straits Times, 26 Apr 2015 Pull-cord alert alarm system (AAS) @ home 5

Technology Pilots & Services 6

Technology-enabled community care “Can non-intrusive technologies be used to better enable community care for me to age-in-place?” - Mr Tan, 78yo, living alone Key Takeaways Supportive Accessibility & Data-driven care ecosystem beyond Unobtrusiveness can improve technology for sustained use wellness Source: The Straits Times, 12 April 2012 *SHINESeniors (Nov 14 – Oct 17) is an SMU-led research project supported by the Ministry of National Development and National 7 Research Foundation under the Land and Liveability National Innovation Challenge (L2NIC) Award No. L2NICCFP1-2013-5.

IoT-Enabled Community Care Ecosystem Assessment of community care needs of Aging-related Policy Enablers elderly living alone Requirements Care findings User-feedback & Refinements Multi-modal data collection Personalized, - In-home unobtrusive As-Needed monitoring Community Care - Survey & ad-hoc observations ADL-based Care models Multi-modal analytics - Wellbeing assessment - Activity/wellbeing-based care Technology Regional Community alert Enablers Care Enablers 8

Understanding elderlies’ needs & wellbeing Social-demographic profile, family support, financial status Physical health, mental health, medication, sleep patterns and quality, activities of daily living Psychosocial Surveys & Regular Social function, overall happiness and wellbeing, liveability, Ground Observations technology Routines and unusual events (hospitalization, faint spells, family visits etc) 9

Physical health profile No of chronic conditions Top 10 physical health conditions (n = 105) 12 Cataract 70.48% 10 High blood pressure 64.76% 8 Hyperlipidemia 60.95% No of Elderly Joint pain, arthritis, rheumatism or 46.67% nerve pain 6 Diabetes 29.52% 4 Heart attack, angina, chest pain 19.05% Parkinsons 13.33% 2 Chronic back pain 13.33% 0 Other fractures 12.38% 1 2 3 4 5 6 7 8 9 10 11 Digestive illnesses 12.38% SHINES LIONS THK 10

Mental health profile AMT Score Geriatric Depression (GDS Score) 70 60 GL@MP – 1 GL@MP – 14 60 50 LIONS - 4 LIONS - 8 50 % of elderly THK - 0 THK - 2 % of elderly 40 40 30 30 20 20 10 10 0 0 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10 11 12 14 AMT score GDS Score SHINES LIONS THK SHINES LIONS THK A score of 6 or less suggests cognitive A score greater than 5 suggests depression • • impairment at the time of testing: 0-3 Severe impairment • 4-6 Moderate impairment • 6 Normal • 11

In-home unobtrusive monitoring Le Lege gend: Motion Sensor Gateway Help button Sensorized Door medication box Contact 12

Activity-based alert and care Refinements Help request Community Inferred medication Care Model Data analysis & anomaly detection Anomaly-triggered Activity-based alert Care Personalized evaluation Zonal activity level Overall activity Care Execution Going out Help button @ home level @home alert 13

Example: Prolonged Inactivity @ home A period of prolonged inactivity at home can indicate trouble for the elderly resident When inactivity exceeds a threshold, trigger an alert to caregivers Challenge: How to set the right alert threshold for different elderly with different daily routines 14

Example: Prolonged Inactivity @ home High blood pressure, Early 80s diabetes and high cholesterol Aunty Tan Aunty Chan Goes out frequently, daily exercise Stays home mostly routine Frail, pain in legs, fall history Generally fit with no fall history Socializes frequently with family and Socializes infrequently, with few visitors neighbours, with regular visitors 15

Survey on inferred medication ~50% of elderly ~60% of elderly store medication in do not pack their medication plastic bags or containers 16

IoT solution to detect medication non-adherence Sensorized medication box Detection of medication non-adherence Caregiver group alert & intervention To fit existing habits 17

Intervention improves adherence Improved medication adherence after intervention in Sep 2016 (medication reconciliation) 18

Wellness-based notification and care Refinements Activity level @ Activity level @ Wellbeing indices bedroom kitchen • Loneliness/ ss/Soci cial al Multi-modal Isolati tion on • Frailty Community Data analysis • Depression Care Model • … Overall activity Activity level @ Going out level @home bathroom Wellness-based Care Personalized evaluation Care Execution Poor / worsening wellness level? 19

Detecting loneliness in elderly living alone Loneliness Depression (Geriatric depression scale) Social loneliness Emotional loneliness Sleep (Pittsburgh sleep quality index) E.g. There is always someone I E.g. I experience a sense of can talk to about my day-to-day emptiness Correlated problems Cognition (Abbreviated mental test) 11% 11% 7% 7% 13% 3% 6 to 10 5 to 10 13% 3% 38% 38% 11 to 15 11 to 15 24 24% 16 to 20 16 to 20 IADLs 21 to 25 27% 27 40% 0% 21 to 25 (Instrumental activities of daily living) 26 to 30 27 27% 20 Prevalence among the Marine Parade elderly sample

Detecting social isolation of elderly Social Sensor-derived Emotional Social Social Isolation feature loneliness loneliness network Score Average daily away -0.22 -0.38* 0.31* -0.42* duration (0.144) (0.011) (0.037) (0.005*) 0.13 -0.10 -0.07 0.08 Away count (0.392) (0.503) (0.656) (0.606) -0.08 0.32* -0.26 -0.05 Napping duration (0.597) (0.038) (0.101) (0.777) Night time sleep -0.12 0.24 -0.14 -0.16) duration (0.448) (0.133) (0.373) (0.297 Average time spent 0.31* -0.01 -0.23 0.17 AWAY DURATION, NAPPING DURATION and in the living room (0.049) (0.973) (0.149) (0.292) TIME SPENT IN THE LIVING ROOM are correlated with social isolation dimensions -0.11 0.03 0.03 0.10 Kitchen activity (0.48) (0.854) (0.852) (0.508) P values are in parenthesis *** p < 0.001, ** p < 0.01, * p < 0.05 21

Impact of SHINESeniors on Ageing-in-Place 52 Help requests Medication adherence timely assistance to 8 for 24 elderly elderly in 13 cases personalized intervention 80 elderly improved adherence in 2 elderly 2017 beneficiaries 2016 3 care partners, 14 elderly @ risk 2 estates, of social isolation prolonged inactivity 3 Govt partners 2015 personalized Detected in 17 elderly intervention achieved 1 elderly found unwell reduced isolation and warded in time! 22

Thank you for your attention Hwee-Pink TAN, Ph.D. hptan@smu.edu.sg icity.smu.edu.sg