KUAC

Inspiration

Our inspiration for our project derived from Amazon Alexa and GPTor, and wanted to make it ourselves.

What it does

The microphone takes voice input, then transcribes the input into chatgpt, where it will then generate a response. This response is both text and can be text-to-speech. After that, the response is played back using the desired output path. The robot also dances when you greet it with a "hello".

How we built it

The body for the robot was built using laser cut plywood material. Servo motors were used to move the arms of the robot. A raspberry pi 4b was used to run the scripts. OpenAi API was used to generate the responses. Pyttsx3 library was used to read the response along with the SpeechRecogntion library to take input and read it. GPIOZero library was used to control the motors with our code.

Challenges / Troubleshooting

Many different implementations for the bot were prototypes on paper making it trouble to construct a workflow. Disagreement on preferred models, faulty implementation of electronic components and scripts. The most challenging issue was that there was not a suitable microphone to connect the raspberry pi.

Accomplishments that we're proud of

We realized we were making many errors and took time to step away from the maker space to cooperate and think of an idea before stepping back into it. Through stress and tiredness our team came together, refined our workflow, composed a solid final draft, and worked as a team to complete our project.

What we learned

Uses of various sensors. Component attachment and debugging procedure for the raspberry pi. Coding to the raspberry pi terminal, and CAD models.

What's next for KUAC

Making the robot move and have a password system so unregistered users cannot give requests to it.

Built With

• python
• pyttsx3
• raspberry-pi
• speechrecognition