Speculative design and ubiquitous computing for a non-coder

Palette Plate is an integrated kitchen tool featuring a projector, camera, and lighting designed to spark creativity in plating. It uses AI technologies like OpenAI, Foodvisor, and YOLOv8 to analyze food and plates, then projects plating guides directly onto the kitchen counter.

Here’s how it works:

1. Food Recognition: Place ingredients on the counter, and the built-in camera scans and identifies the size and shape of the food using computer vision.
2. AI Analysis: The system processes the food data, providing nutritional information and plating suggestions based on the type and shape of the ingredients.
3. Creative Plating Suggestions: The AI generates plating designs using tools like DALL·E, offering creative ideas on how to arrange the food.
4. Projection Guidance: The system projects these suggestions onto the counter, showing exact placement and boundaries for the ingredients.
5. Plating Execution: Users follow the projections to create a precise and aesthetically pleasing dish, enhancing both creativity and efficiency in the kitchen.

Palette Plate seamlessly blends tech with culinary artistry, empowering users to plate like a pro while advancing innovation in the kitchen.

During development, our team faced several challenges, particularly with hardware design and projection interaction. On the hardware side, we had to ensure that the projector, camera, and lighting components worked harmoniously within a compact device. This required careful engineering and multiple prototype tests, leading us to refine the material choices and layout for optimal functionality and user ease.

For the projection experience, our main hurdle was ensuring accuracy and creating a user-friendly interface. We realized that the food recognition and plating suggestions could take over 10 seconds, which might cause user frustration. To address this, we incorporated loading animations into the software.

So, let me be real for a second—coding was not my thing. I had zero experience when I started, and honestly, it felt like I was learning a new language (or five). But, after living on StackOverflow and basically having ChatGPT as my coding BFF, I made it work! Every bug felt like a mountain, but nothing beats that moment when you finally crack the code and everything just works.