X
Innovation

Domino's is tapping Nvidia GPUs to ramp up AI-powered pizza delivery

Nvidia is highlighting the use case at the National Retail Federation's annual conference in New York this week, demonstrating the power of AI in any kind of brick-and-mortar storefront.
Written by Stephanie Condon, Senior Writer

Pizza delivery has become a cutting-edge business: Pizza Hut, for example, recruited Pepper the Robot in 2016 to take customer orders. Little Caesars has patented a pizza-making robot. Domino's, meanwhile, has teamed up with Ford to deploy self-driving delivery vehicles, and it's conducted drone deliveries

To take its technical operations to the next level, Domino's is leveraging Nvidia GPUs to accelerate and improve its AI-powered applications. 

Domino's "has grown our data science team exponentially over the last few years, driven by the impact we've had on translating analytics insights into action items for the business team," Zack Fragoso, a data science and AI manager at the pizza company, said in a blog post published by Nvidia.  

Nvidia is highlighting the use case at the National Retail Federation's annual conference in New York this week, demonstrating the power of AI in any kind of brick-and-mortar storefront. 

Domino's launched its most high-profile AI project last year during the Super Bowl. The "Points for Pie" app that invited users to submit a photo of any kind of pizza in exchange for customer loyalty points. Using more than 5,000 images, the company trained a machine learning model to classify pizza images with an Nvidia DGX system equipped with eight V100 Tensor Core GPUs. Domino's created a unique dataset with all of the pictures submitted via the app, which it considers a strategic corporate asset. 

The company has also used GPUs to boost the accuracy of predicting when an order will be ready. While it sounds like a simple predict, the variables involved include how many managers and employees are working, the number and complexity of orders in the pipeline and traffic conditions. With a DGX server, Domino's improved the accuracy rate from 75 percent to 95 percent. 

Now, Domino's is using a bank of Nvidia Turing T4 GPUs to accelerate AI inferencing for tasks that involve a range of real-time predictions. For instance, their data science team is exploring computer vision applications both inside and outside the store to improve pizza carryout experience.

The Domino's case study demonstrates how retailers can use AI to create operational efficiencies and improve customer experience, but the benefits go beyond that. There's the potential, for instance, to improve logistics, optimize merchandising decisions or measure store traffic. 

Editorial standards