AI Business is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 3099067.

Retail & Supply Chain

AI Summit London: How Tesco used AI to help deal with lockdown demand

by Louis Stone, Reporter
Article Image

Tesco’s CTO talks about the steps taken to confront the COVID crisis

The pandemic has upended so much of modern life, transforming where we work and who we meet, making us seek to minimize interactions with other human beings.

Among the rituals most impacted by COVID-19 is how we eat; many retailers were struggling to cope as main streets were left deserted and everyone suddenly flocked online.

For British grocery giant Tesco, the world’s third-largest retailer, the sudden shift presented a significant challenge. The solution, CTO Guus Dekkers told the AI Summit London, lay in a combination of physical changes, smart use of big data, and the latest in artificial intelligence.

When the world changed

“We've seen probably more change in the last handful of months than we've seen in the last couple of years,” Dekkers said. “We all as an industry have been forced to adapt and basically to find ways of how we make sure we can continue serving the nation and make sure that there is access to food for everyone.”

At the beginning of the COVID-19 outbreak, when customers were still rushing to stores to buy long-lasting foods and, for some reason, reams of toilet paper, keeping the shelves full required some quick thinking.

“I think it surprised us all in the way our population reacted,” Dekkers said. “But of course, we have to react. From a tech perspective, one effort was trying to find ways how we can help with availability and make sure that there is an equal distribution of the available items to our customers. So, we implemented a lot of nudging and hinting, to make sure there was decent access to the items which have been available.”

With the demand spike, the company was forced to ask itself whether it could step up in optimizing its supply chain, Dekkers said. “Just to put it in perspective, the demand drove a requirement for us to set up, basically overnight, further warehouses and distribution centers.”

In the UK, the company offers customers time slots, when they can request deliveries of goods, primarily food. Before the pandemic, Tesco offered roughly 500,000-600,000 slots a week, which was able to meet the demand at the time. "And we have grown this to more than double, at the moment it’s hovering around 1.3-1.4 million a week, which we are offering to our customers,” Dekkers said.

Some of this has been made possible thanks to a major hiring push, with the company adding some 30,000 temporary staff. This in itself has been a considerable technical challenge. “How do you onboard them? How do you make sure by the end of the month, you're capable of paying an additional 30,000 new colleagues? All of these have been massive challenges which we as a team picked up and have found answers to in order to make sure that we could react to these challenges,” Dekkers said.

The company also doubled down on its AI modeling, continually retraining its delivery route optimization algorithms on the latest data. “And that means that on a daily basis we run over a billion iterations in which we basically try to optimize our overall delivery schemes and our overall delivery routes,” Dekkers said. He added that the company has been able to decrease total mileage by 7-8 percent, meaning less fuel consumption and CO2 emissions.

“But equally, something which has helped us very much around the COVID crisis is how we can increase our van utilization and make sure that, from a cost perspective, we do the right thing,” he added.

This work joined other optimization efforts, as the company tried to squeeze efficiency out of the assets it already had: “It’s all built around how can we use our existing capacities,” Dekkers said. “How can we make sure that people are picking and realizing these orders, how can we make sure that they can continue doing their job in a safe manner, and COVID-compatible manner, whilst at the same time delivering or reacting to this tremendous growth and demand which we have seen?

“For us, it all translates into: how do we optimize the supply chain? How do you get better predictability on what's going to happen? How do we change things like routing and picking algorithms overnight, in order to make sure that we use our capacity?”

Now that we’re past the first wave and many are returning – albeit, sometimes, begrudgingly – to offices, it’s not clear what the future of commerce will hold. Dekkers is certain that for Tesco, the years ahead will see growing reliance on online orders – especially with the pandemic helping change brand perceptions of the traditionally brick-and-mortar retailer.

“Many of the changes which have been triggered in recent months are not going to go away,” Dekkers said. “We [must] continue creating further capabilities in the online space.”

This will require more physical presence for distribution, but ideally without taking on too much additional land during these uncertain times. Tesco’s solution is to use excess space at the company’s largest stores and convert them into an AI-powered “automated warehouse solution which we call an urban fulfilment center,” Dekkers said. “It is substantially smaller than any traditional automated warehouse you might know.

“Basically we can do this in a sub-part of a larger supermarket, and through this automation we are able to increase both the volume of orders we can deliver from such a store and do it more cost-effectively,” he said. It should also enable new capabilities like same day delivery, or even delivery in the next hour, and enhanced ‘click and collect’ services.

The first urban fulfilment center is set to go live soon, with two more planned this year, and around 10 for 2021.

“It’s a substantial investment in making sure that our capabilities to deliver in the online space continues to grow,” Dekkers said.

Practitioner Portal - for AI practitioners

Story

MLOps startup Verta gets $10m in funding, launches first product

9/1/2020

The company plans to commercialize open source ModelDB project, developed by CEO Manasi Vartak

Story

AI and analytics services: Capabilities and costs

8/27/2020

Which skills do you need in your team? What are the costs for running the service? How can you optimize them? These are three key questions when setting-up and running an AI and analytics service.

Practitioner Portal

EBooks

More EBooks

Upcoming Webinars

Archived Webinars

More Webinars
AI Knowledge Hub

Experts in AI

Partner Perspectives

content from our sponsors

Research Reports

More Research Reports

Infographics

Smart Building AI

Infographics archive

Newsletter Sign Up


Sign Up