Eye on the prize – how AI is perfecting your pizza
An AI integration that helps ensure pizzas are made perfectly is now over a year old. So how successful has it been?
Once upon a time, you might have opened your pizza box and found that “hold the pineapple” or “extra olives” hadn’t translated very well. In fact, according to Domino’s management, “my pizza doesn’t look like it should” was their most common customer complaint.
But one year, 820 stores across Australasia, and 50 million pizzas later, Domino’s says its DOM Pizza Checker – a smart scanner that sits above the pizza cutting bench and double-checks that all the toppings and sauces are successfully in place – has changed that.
“This is the ultimate customer experience,” begins Domino’s Chief Operating Officer Michael Gillespie. “It provides customers with complete transparency from the moment they order to the moment their meal is delivered.”
So how does it work? The technology in the tool can recognise, analyse and grade pizzas based on pizza type, correct toppings and distribution. It does this by capturing an image of the pizza and using AI to compare what it sees with a large dataset of correct pizzas, making a quick judgement on accuracy and quality. Then, the pizza is given an overall rating out of 100. If it falls outside acceptable margins, it’s flagged as a fail, and the pizza is remade. The tool, says Gillespie, ensures “there are no more surprises when it comes to pizza quality and consistency”.
Customers can go on the ride, too – tracking their pizza through the company’s website. “They can have complete visibility over their meal from the moment they order,” says Gillespie. That includes everything from where it is in the making process, to whether it has passed the QA test, and pizza lovers can even check what their meal looks like when it comes out of the oven.
“We worked closely with our technology partner, Dragontail Systems, for more than two years to bring this innovation to life,” Gillespie continues, “all the way from idea conception to execution. This involved everything from developing and installing the physical hardware, to teaching the technology to recognise different pizza types and toppings – through literally thousands of pizzas – and to understand what an ‘approved’ pizza looks like, versus a pizza requiring a remake.”
That’s all very well, but is the tool helping staff make a better pizza to begin with? Apparently so.
Data gathered in the back end is used for internal reporting purposes. Domino’s staff can monitor the stats and images collected by the pizza checker and identify areas for improvement or further training.
Gillespie says it’s an example of AI working hand-in-hand with a human labour force to help make better pizzas, and that it’s “an invaluable training tool, ensuring that even in busy periods, our customers are getting the highest quality pizzas possible.”
So far, Domino’s is the only quick-service restaurant, or QSR, in Australia or New Zealand sharing real-time images of the food that’s being produced. And, Gillespie says, the dividends are in their satisfied customers. “The DOM Pizza Checker has provided a platform for transparency – it’s Domino’s solution to this customer tension point.”