WanderGuardian: A Mobile App to Mitigate Spread of COVID-19 from Wanderers
A Project Funded by Alliance of Chinese Americans San Diego
Edmund Sumpena
Background
Wandering among people with dementia or developmentally disabled children is a serious problem; it can lead to serious injury or fatality. This group includes patients of Alzheimer’s disease, Parkinson’s disease, autism, Down syndrome, or other neurological disorders involving memory loss. The problem is further magnified by the COVID-19 pandemic, where wandering can worsen the spread of the virus.
According to the Los Angeles Times*, the eldercare facility has become a tragic focal point of the COVID-19 outbreak; nearly half of all COVID-19 deaths in California are linked to the eldercare facility. To minimize the risk of contracting the disease, many family members of the residents opted to check them out of the facility and care for them at home.
A home environment, unlike memory care facilities, lack the necessary elderly care technology, resulting in an elevated risk of outdoor wandering. By being outdoors, wanderers are not socially distancing, putting themselves and surrounding people at risk of developing or spreading COVID-19, as some may be asymptomatic.
To mitigate the wandering-related spread of the virus, I have developed a cost-effective, indoor-compatible, high-sensitivity, and privacy-protecting mobile Geofencing technology to complement existing solutions. The app, WanderGuardian, serves as an early warning system. A caretaker or family member can set up a predefined safe zone and remotely determine when a patient is within this area. If the patient wanders outside the safe zone, the app will notify the caretaker within seconds. This way, caretakers or family members can prevent possible accidents and limit the spread of COVID-19 by wanderers in the local community.
*Los Angeles Times Article: https://www.latimes.com/california/story/2020-05-08/california-coronavirus-deaths-nearly-half-linked-to-elder-care-facilities?utm_source=sfmc_100035609&utm_medium=email&utm_campaign=25259+Today%27s+Headlines+5%2f28%2f2020&utm_term=https%3a%2f%2fwww.latimes.com%2fcalifornia%2fstory%2f2020-05-08%2fcalifornia-coronavirus-deaths-nearly-half-linked-to-elder-care-facilities&utm_id=7469&sfmc_id=299825
Wandering among people with dementia or developmentally disabled children is a serious problem; it can lead to serious injury or fatality. This group includes patients of Alzheimer’s disease, Parkinson’s disease, autism, Down syndrome, or other neurological disorders involving memory loss. The problem is further magnified by the COVID-19 pandemic, where wandering can worsen the spread of the virus.
According to the Los Angeles Times*, the eldercare facility has become a tragic focal point of the COVID-19 outbreak; nearly half of all COVID-19 deaths in California are linked to the eldercare facility. To minimize the risk of contracting the disease, many family members of the residents opted to check them out of the facility and care for them at home.
A home environment, unlike memory care facilities, lack the necessary elderly care technology, resulting in an elevated risk of outdoor wandering. By being outdoors, wanderers are not socially distancing, putting themselves and surrounding people at risk of developing or spreading COVID-19, as some may be asymptomatic.
To mitigate the wandering-related spread of the virus, I have developed a cost-effective, indoor-compatible, high-sensitivity, and privacy-protecting mobile Geofencing technology to complement existing solutions. The app, WanderGuardian, serves as an early warning system. A caretaker or family member can set up a predefined safe zone and remotely determine when a patient is within this area. If the patient wanders outside the safe zone, the app will notify the caretaker within seconds. This way, caretakers or family members can prevent possible accidents and limit the spread of COVID-19 by wanderers in the local community.
*Los Angeles Times Article: https://www.latimes.com/california/story/2020-05-08/california-coronavirus-deaths-nearly-half-linked-to-elder-care-facilities?utm_source=sfmc_100035609&utm_medium=email&utm_campaign=25259+Today%27s+Headlines+5%2f28%2f2020&utm_term=https%3a%2f%2fwww.latimes.com%2fcalifornia%2fstory%2f2020-05-08%2fcalifornia-coronavirus-deaths-nearly-half-linked-to-elder-care-facilities&utm_id=7469&sfmc_id=299825
Goal
This mobile app’s goal is to limit wandering-related transmission of COVID-19 as well as to raise public awareness about the potential dangers of wandering. I also hope this technology will ease the burden of families who are caring for dementia patients, as well as developmentally disabled children at home during pandemic. For my next phase, I will be implementing Bluetooth beacons, which cost $20 each. In the situation when a wanderer forgets to bring their phone, these pocket-sized devices will allow WanderGuardian to detect when the user is in the vicinity of their mobile device. If the wanderer is too far away from their phone, the caregiver is alerted. It will be complete by end of June.
This mobile app’s goal is to limit wandering-related transmission of COVID-19 as well as to raise public awareness about the potential dangers of wandering. I also hope this technology will ease the burden of families who are caring for dementia patients, as well as developmentally disabled children at home during pandemic. For my next phase, I will be implementing Bluetooth beacons, which cost $20 each. In the situation when a wanderer forgets to bring their phone, these pocket-sized devices will allow WanderGuardian to detect when the user is in the vicinity of their mobile device. If the wanderer is too far away from their phone, the caregiver is alerted. It will be complete by end of June.
Method
1. Interviewed elderly care facilities to better understand the challenges of wandering and how COVID-19 has impacted them.
2. Researched machine learning technique for safe zone detection. Built a test app to collect sensor data and test the machine learning algorithm. Continued to improve the algorithm’s accuracy by adjusting training time, adding calibration features, editing constants and calculations.
3. Improved user interface (UI) and debugged.
4. Tested the application: collected data on the algorithm’s accuracy, improved user-friendliness, and received feedback.
1. Interviewed elderly care facilities to better understand the challenges of wandering and how COVID-19 has impacted them.
2. Researched machine learning technique for safe zone detection. Built a test app to collect sensor data and test the machine learning algorithm. Continued to improve the algorithm’s accuracy by adjusting training time, adding calibration features, editing constants and calculations.
3. Improved user interface (UI) and debugged.
4. Tested the application: collected data on the algorithm’s accuracy, improved user-friendliness, and received feedback.
Update
Progress Update
(October 3, 2020)
WanderGuardian is designed with a privacy protecting, sensitive, cost effective, indoor compatible, and adaptable machine learning algorithm. A caretaker sets up a predefined safe zone and remotely determines whether a patient is within the zone. If the patient wanders outside the area, the app will notify the caretaker in a timely manner. Therefore, WanderGuardian acts as an early-warning system to prevent the possibility of wandering.
I have also implemented wearable Bluetooth beacons, which cost an average of $20 each, into the app. In the situation where a wanderer forgets to bring his/her phone, WanderGuardian will detect this pocket-sized device and determine whether the user is in the vicinity. If the wanderer is far away from the phone, the app will alert the caregiver.
Winner of Microsoft Imagine Cup Junior 2020
I would like to share with ACA that my work on artificial intelligence (AI) in WanderGuardian has won the Microsoft Imagine Cup Junior 2020 award. This is a global AI for Good competition that encourages students to design and create technology to solve social, cultural and environmental problems in the world. Anthony Salcito, Vice President of Microsoft Education, celebrated the winning projects in the Microsoft Education Blog. The announcement was also posted on social media:
Twitter Facebook LinkedIn Instagram
(October 3, 2020)
WanderGuardian is designed with a privacy protecting, sensitive, cost effective, indoor compatible, and adaptable machine learning algorithm. A caretaker sets up a predefined safe zone and remotely determines whether a patient is within the zone. If the patient wanders outside the area, the app will notify the caretaker in a timely manner. Therefore, WanderGuardian acts as an early-warning system to prevent the possibility of wandering.
I have also implemented wearable Bluetooth beacons, which cost an average of $20 each, into the app. In the situation where a wanderer forgets to bring his/her phone, WanderGuardian will detect this pocket-sized device and determine whether the user is in the vicinity. If the wanderer is far away from the phone, the app will alert the caregiver.
Winner of Microsoft Imagine Cup Junior 2020
I would like to share with ACA that my work on artificial intelligence (AI) in WanderGuardian has won the Microsoft Imagine Cup Junior 2020 award. This is a global AI for Good competition that encourages students to design and create technology to solve social, cultural and environmental problems in the world. Anthony Salcito, Vice President of Microsoft Education, celebrated the winning projects in the Microsoft Education Blog. The announcement was also posted on social media:
Twitter Facebook LinkedIn Instagram