Yobo – The Stir Robo

Millions of home cooks spend hours stirring food so it doesn't burn. Yobo combines a motorized stirrer, live monitoring dashboard, and AI learning system for smarter slow cooking.

Comment

Inspiration

We both love traditional Indian homemade desserts. We have seen our mothers stirring a pot for 4–5 hours so the food would not stick or burn. We always wanted to have a tool that could do that for us. We are software engineers, but we did not have any hardware knowledge. We wanted to reduce that repetitive work without changing how the food is cooked and integrate sensors data along with Machine learning in cooking.

What it does

Yobo takes over the repetitive stirring needed for slow-cooked foods. It keeps the food moving, lets you watch the pot live from your phone, tracks cooking time and energy use, automatically controls the exhaust fan, and creates a timelapse and summary after each cooking session. It is also being trained to recognize the pot, ingredients, paddle movement, and different stages of cooking.

How we built it

We started with a basic motorized stirrer and improved it through several prototypes. The current version combines an aluminum frame, geared motor, adjustable wooden paddle, camera, smart energy plug, sensors, and a dashboard built with help from ChatGPT, Image generation, Codex and GPT-5.6.

Challenges we ran into

At first, we did not know how to make it, what materials to buy, or where to buy them. ChatGPT helped us with the research, and image generation helped us visualize the items before we even bought them. Thick food created more motor load than expected. We also faced paddle-scraping issues, pot movement, shaft alignment, heat near electronics, camera reliability, and difficulty correctly detecting when cooking sessions started and ended. For the dashboard, we also faced the challenge of deciding which technologies to use.

Accomplishments that we're proud of

Yobo has completed real, multi-hour cooking sessions multiple times. We are proud that the hardware, live monitoring, energy tracking, exhaust-fan automation, timelapses, and session summaries now work together as one system and we don't need to sit on pot constantly stiring.

What we learned

We learned that hardware must be tested under real cooking conditions. Codex helped us quickly turn each test result into software improvements and new iterations.

What's next for Yobo – The Stir Robo

Next, we want Yobo to detect paddle movement, boiling, ingredients, and cooking stages. Eventually, it will be able to use the camera to detect what food we are cooking and recommend or automatically adjust the stirring speed and stove heat.

Built With

Share this project:

Updates