This week, I focused on making WebSockets communications between our WebApp and the RPi work. When we met up on Thursday afternoon, Mahlet and Jeffrey helped to set up the RPi (registering the device, connecting it to WiFi, etc.). Once we were able to download VSCode on the RPi, I coded up a short script to test if communications were able to happen. I wrote a simple script in JavaScript on the RPi, and then wrote a similar one with some extra UI features on the WebApp and tested it out. Theoretically, when I clicked a button on the WebApp, the RPi should receive it and print out a message. Initially, this wasn’t working due to the RPi and the WebApp being on different ports. There was a CORS(Cross-Origin Resource Sharing) error, due to the WebApp trying to send a request to a different domain than the server that was hosting it and so to debug this, I included some CORS settings on the RPi side to allow the WebApp to send a request. This worked, and the RPi was able to display a message when a button on the WebApp was clicked.
On the WebApp:
On the RPi:
I also spent quite some time on trying to incorporate TTS on the WebApp itself this week. Unfortunately, the pyttsx3 library that we were trying to use seems to not work well with our website. After coding up some simple logic to use the TTS function in the library when a user input is received, me and Mahlet tried testing it to see if it was successful. When we first input some text into the textbox and click the read button, it works well and the laptop speakers play the correct audio with little to no delay. However, when we try to send more text again, we get the error “run loop has already started”, which indicates that the previous text to speech command queued had not finished. We were confused and spent quite some time trying to debug this by looking up solutions online that other users who have encountered this issue tried and it did not work for us. We looked through the documentation for the TTS library itself and tried out various functions, but nothing seemed to work. Thus, me and Mahlet are looking into using other TTS libraries to see if we can find a solution to this. I am considering using gTTS (Google Text to Speech), which is not as ideal as pyttsx3 because it requires an internet connection, but should be well-documented enough to reduce the chances of it not working as well.
In the next week, I’ll be working on:
- Building the robot with my team
- Researching different solutions for TTS with Mahlet
- RPS game function on the WebApp with WebSockets