General Usage - xxyypp/IBM_Dementia GitHub Wiki
About
This page explains how to use all of the device's current functionalities. Remember that to work like intended, the Dementia Protector needs to be built by exactly following the steps presented on the previous pages.
Instructions
1. Turning on
To turn on the device, simply power the Raspberry Pi using a USB cable. After some delays the right webpages should open (If they don't, make sure that the last setup steps on the "Node-RED" page were followed correctly). You might have to refresh the pages if they don't open properly.
2. Connection
The first time the device is turned on, the Raspberry Pi needs to be connected to the phone. A simple button on the top-right of the Desktop is available for that purpose. Instructions to edit the App for a new phone are available on the Android App page of the wiki.
3. Putting the device on
Place the pulse sensor under the the index finger and the rest of the device on the wrist, like a watch.
4. Talking to Paige
To start a conversation with Paige, say "Hello Page" or press the top button on the screen. Anything else will be ignored by the device.
Ensure that the UI displays mic active
before you start speaking. The speaking
webpage might run into an error, like we said on the "Node-RED" page, if it doesn't start listening again, just refresh the page.
5. Request assistance
If you require direct assistance and don't want to go through the whole conversation process, just press the "HELP" button on either the screen or the phone, or say "help".
6. Checking on the patient from outside
Right now the patient health data is logged in the cloud, which can be accessed in 2 ways:
- Via the Node-Red dashboard which displays the data as time series (beats-per-minute, skin temperature, and skin humidity). This is possible because the Node-Red dashboard is hosted externally via Dataplicity. The user however can change the hosting method.
- Via Dropbox where the above data is stored as a csv file. The csv file can be easily read using excel or other programs which allows for further analysis using statistical or machine learning methods.
7. Playing audio
To hear what the assistant says, either connect an audio device to the Raspberry's jack connection or open the UI by logging onto the IP address of the Raspberry Pi on another device (that has built in audio) that is connected to the same network (also the phone).
8. Exiting the UI after launch
The UI automatically opens on a full screen chromium tab. To exit this page you will have to close chromium (Alt+F4).
You can also switch to the speaking
page by pressing CTRL+SHFT+TAB and refresh it with CTRL+F5.
NOTE: In order to do all the setup you will need to connect a keyboard (or mouse) to the USB entrance of the Device.