aCtFasteR's primary aim is to support our elderly living alone. They are most vulnerable to the risk of not receiving timely medical assistance after an incident has occurred, since there is no one around to help them and they may not be physically abled to notify anyone for help. This causes the most harmless of falls to become fatal in the event no aid is rendered for a few days. In the unfortunate scenario where our elderly has suffered a mishap, there will be inactivity in the house for an extended period of time. Through the use of proximity sensors and the Internet of Things technology, we aim to integrate aCtFasteR to the MyResponders App, so that we can tap on nearby CFRs to respond quickly in the event of inactivity. Detecting an elderly in distress earlier can save another life.
Note: Our solution works on the assumption that an elderly will be moving around the house to carry out their daily activities, such as going to the toilet, having a meal, taking a nap etc. This is a valid assumption for an elderly living alone as they will have to carry out all household chores independently.
Privacy is a prominent issue with regards to the use of technology in today's world. Though visual and voice surveillance offer a more comprehensive monitoring of the elderly's well-being, it is often too intrusive and results in both privacy and security concerns, especially if the data is not stored properly. Hence, our goal is to implement an innovative monitoring system without infringing on one's privacy at all. To achieve this, we have decided to only make use of proximity sensors to track the activity in one's household.
There is a wide range of scenarios that can result in inactivity in the house, for example the elderly could have been spending time outside of the house or they could have been asleep. Our solution has thus implemented 3 measures to reduce the rates of false positives.
- With a database of the elderly's details, a call will be made to the elderly to confirm his/her well-being. If the elderly cannot be reached, a call will be made to the elderly's next-of-kin if any.
- The limit for the duration of inactivity will be extended at night to account for the time that the elderly is asleep. With these measures, our team aims to reduce the chances of false detection by aCtFasteR. Note: The exact time whereby the duration of inactivity is extended may be customised depending on one's sleep habits.
- A switch will be installed at the front door to turn off the system as the elderly is leaving their home for extended period of time. We hope that this will reduce false alarms caused by our solution.
In the event of a suspected incident, it is crucial that the elderly receives timely medical assistance as soon as possible. The process to deploy a CFR to the elderly's location has to be expedited. Hence, we have to propose a trigger of an appropriate sensitivity, so that we can render aid to the elderly in distress early. In the event that we are unable to verify an elderly's well-being within 15 minutes (the esimated time for a few phone calls to verify the elderly's well-being), CFRs within the area will be activated to check in with the elderly. This is a fail-safe measure.
- When a user signs up for aCtFasteR, they will be required to enter their details, which includes their phone number, address and next-of-kin details, if any. These data will be stored in a database and will only be disclosed to a CFR in the event of an activation.
- Proximity sensors will be installed in the house at multiple doorways. For example:
Using the proximity sensors to detect motion, data will be sent out to a cloud whenever the elderly walks past the sensor. This data is processed as activity. In the event the elderly has to leave the house for an extended period of time, a button will be installed in the house at the main door to allow the elderly to halt the detection system.
When the proximity sensor near the main door detects motion after the elderly returns home, the system will be reactivated. To prevent the proximity sensor from reactivating the system while the elderly is in the midst of leaving the house, the detection system will only be turned off 1 minute* after the button is pressed. - The system is triggered by a timer which will start counting down when there is inactivity. Whenever an activity is detected in daytime (7am to 7pm), the timer will be resetted to six hours* and the countdown restarts. The rationale for six hours as our threshold is that as long as the elderly is at home, he/she will have to go through their daily routine and move at least twice during the day. During the night (7pm to 7am), the timer will be resetted to twelve hours* to account for sleep. This means that as long as there is activity, the countdown will never end and the alert will not be triggered. *Note: Timing given is arbituary and subjected to changes.
- In the event of an incident and the elderly is unable to move to request for help, the proximity sensors will stop detecting any activity and hence no data will be sent out to the cloud system. Since the countdown will not be resetted, it will eventually end and the system will be triggered to attempt to contact the elderly or their next-of-kin through a phone call.
- If the elderly was verified to be safe, CFRs will not be alerted and the system will be reset.
- Else, an alert will be sent out via the MyResponders App to nearby CFRs. Along with the alert, the address of the elderly will be sent to the CFRs from the database.
- Upon arriving at the elderly's apartment, the CFR will access the situation and determine if there is a need to contact emergency services. This will help with early intervention as compared to if the elderly in distress was waiting to be discovered.
For the purpose of our demonstration:
- An IoT application will be used on an android phone to simulate the proximity sensor. The flow of the data will be represented by a flow on Node-RED.
- The threshold duration will be reduced to seconds for us to test out the system easily in real time.
- Once the app stops sending positive data (ie no movement recorded by proximity sensor) for the threshold duration, an alert message will be sent out which will appear on the dashboard.
- The checking of false alarm will not be depicted in the flow as the process will be running on the MyResponders App and there is no database set up for now. Similarly, the alert message will not be shown on the MyResponders App itself, but rather on the dashboard.
- We envision the use of data analytics to further refine the threshold for periods of inactivity, above which an alert will be sent. Doing so will allow the number of hours set before an alert is sounded to be highly customised according to each individual.
- To further establish the culture of inter-dependency, caretakers and NOK could be given the same access rights as the elderly they are interacting with. This allows them to momentarily overwrite and deactivate the aCtFasteR system through a web-based server in the event the elderly forgot to do so, further reducing the likelihood of false alarms.
- Coloured-lights can be installed on the front doors that lights up if the elderly was suspected to be in distress.