In a new blog released on Microsoft's website, PepsiCo describes how the firm leverages AI to make Cheetos so tasty and light. The secret lies behind an AI solution powered by Microsoft Project Bonsai.
The future for process controls
“This is the future for process controls,” said in the blog Sean Eichenlaub, a senior principal engineer at PepsiCo. “We’re using AI-based automation to improve the consistency of our products.”
The goal of Microsoft Project Bonsai was to more efficiently and consistently manufacture ideal Cheetos while reducing waste. To do this they built a computer vision system that continually monitors Cheetos attributes such as density and length and makes adjustments to bring the product within specifications.
The approach proved quite successful allowing PepsiCo to reduce the time it takes to tackle inconsistencies in the product line and freeing up operators to focus on the parts of the line that require human adjustments.
“Project Bonsai allows us to utilize technology for the things it does best and our employees for the things they do best,” said Kevin Lin, associate principal engineer at PepsiCo. “The two complement each other and so far, we’ve seen great results.”
The AI project can even recognize and alert operators if the product drifts out of spec. It then proceeds to indicate the attributes that are not ideal and makes recommendations to correct them. In order to fix these issues, the operator needs to simply push a button.
The Project Bonsai solution monitors the product almost continuously, ensuring it is alerted as soon as the product strays outside a defined range. That way PepsiCo guarantees its Cheetos remain just perfect.
Now, the firm hopes to make the AI solution run autonomously. By empowering Project Bonsai to monitor and adjust the process of making Cheetos continuously and independently, the company hopes to maintain Cheetos quality and produce more throughput.