How Pepsi Has Baked AI into Its DNA


Artificial intelligence is big business these days. Thankfully, no AI companies named Skynet have emerged yet, but the ability of computers to learn and perform complex data analysis is growing steadily. We may still be a while away from true cognitive machines, but AI is already showing great potential.

Twenty-three percent of businesses have already incorporated AI technology into their processes and/or product and service offerings, with 54% of executives at those companies stating that the technology has helped them to improve productivity. Meanwhile, 61% of business executives with an innovation strategy say they are using AI to identify opportunities in data that would otherwise be missed. It's predicted that the AI market will grow to a $190 billion industry by 2025, covering everything from data analysis to customer experience.

As one of the world's biggest food and beverage brands, Pepsi is regularly in the vanguard of innovation and has been investing heavily in artificial intelligence technology to boost almost every corner of its business.

Pepsi

The PepsiCo organization sells products in more than 200 countries and brought in $64.7 billion in annual revenue last year, meaning it constantly has thousands of plates spinning at once. Thankfully, it has the financial clout and infrastructure to implement the technology it needs to make those plates twirl a little more smoothly.



One customer-facing application of AI technology can be found making its way around the University of the Pacific. The so-called "Snack-Bot" is an AI-powered delivery robot from which students can order a range of healthy snacks — sparkling water, healthy potato chips, etc. They can have them delivered — with no additional fee — to more than 50 locations around the campus. The technology in the robot lets it navigate at all times of the day and in all weather conditions and has a 20-mile range on a single charge.

The Snack Bot was created based on feedback from university students and aims to help them make healthy snack choices by making the service extra convenient. The response from students has been great, and Pepsi has three to five of the machines operating at any one time to keep up with demand.

Manufacturing

Another place where Pepsi is innovating using AI technology is in the manufacture of its famous Frito-Lay's potato chip subsidiary. The brand wanted to develop a method to help it determine the texture of its chips without handling or destroying them.

To this end, it developed a new laser sensor array that could bounce beams of light off the finished chips and then use AI technology to process the sound feedback, which would create a reading of the snack food's texture.

"One of my first projects at the company involved building systems that could sense the texture of chips without destroying them," said Senior Research and Development Engineer at PepsiCo, Shahmeer Mirza. "It's a system that hits the chip with lasers, listens to the sound coming off them, and then uses that data to correlate the sound into texture. I realized I could combine the principles of machine learning with computer vision technology to develop more advanced sensors and process control schemes."

Building on the success of this innovation, PepsiCo began work on a new type of machine learning application which could determine the size and weight of potatoes. Up until now, the company had been gauging potato numbers based on a vision system, but the team quickly realized that machine learning could be applied to make the process more streamlined and accurate, using visual data captured by cameras to determine a mass flow estimation of the product moving along the conveyor.

However, the brand wasn't ready to stop there and is working on several different AI applications for the Frito-Lay's manufacturing process.

"For example, we have a vision system that looks at every potato going through our peeling process — where we abrasively peel our potatoes," continued Mirza. "I wrote an algorithm that can tell us what our percent peel is; that is, how peeled a potato is after going through the process versus how unpeeled it is. With this information, we can optimize our peeling, so we don't over-abrate the potato and peel away some of its flesh."

Final Thoughts

We can easily see how PepsiCo has successfully baked AI technology into the DNA of its brands. With many companies still demonstrating a significant lack of AI maturity, it takes the biggest brands to lead the way and demonstrate how the technology can be used. We'll have to wait and see how many others follow Pepsi's lead in the future.


AI technology is set to be a hot topic at Digital Food and Beverage 2020, taking place in July at the Hilton Austin TX.

Download the agenda today for more information and insights.



WBR Insights is the custom research division of Worldwide Business Research (WBR). Our mission is to help inform and educate key stakeholders with research-based whitepapers, webinars, digital summits, and other thought-leadership assets while achieving our clients' strategic goals.



Return to Blog