AI Nose

Have you left your house in a jiffy to wonder, a while later, if you’d turned off the gas stove? ✋

Some challenges are universal. Especially those from the kitchen.

In my home, when I prepare the traditional appam, I cover the batter and set it aside to ferment for 8 hours or so. After fermenting the batter will have a fermented yeasty smell. Aside from the look of the batter (risen/ puffed up), the smell is an indicator of readiness.

Halfway across the globe, a PM at #Microsoft faced a similar challenge. He wanted to determine when his sourdough bread starter was ready to bake. To improve his breadmaking skills, he built an #AI Nose- a device that uses AI to tell him when his bread had properly risen. Using a readily available gas sensor and microcontroller, he collected the gas makeup of a smell and used this data to train an AI model.

“The Artificial Nose uses a neural network to correlate the concentration of gases in the air to categories of smell. When connected to an #IoT platform, this nose could be used for a variety of scenarios: it could help to craft a real-time alert system for when foods have spoiled, or for detecting specific gases in the air.”

What other use cases do you think will benefit from this?
E.G. Alert maintenance operator of a gas leak detected on a factory floor or a consumer about spoiled food in their fridge.

Peaceful shopping tops my list. 🙂

Image Credits: Microsoft

Code: https://github.com/kartben/artificial-nose

Demo: https://artificial-nose.azurewebsites.net/en-us/smell1