Deliver America: AI is key to logistics – but do it right
Artificial intelligence was the dominant topic at the recent Deliver America 2025 conference in Las Vegas, but not as a cure-all.
Multiple presentations at Deliver America 2025, held in Las Vegas Oct. 29-30, focused on various aspects of how AI can revolutionize and transform supply chain operations. However, speakers made it clear that retailers and brands need to deploy specific AI solutions that resolve targeted business needs.
Following are overviews of a few key sessions that reinforced the vast potential AI holds for supply chain enterprises, when rolled out to meet clear business objectives.
Fabletics – Win over Gen Z with predictive models and RFID
Opal Portis, VP, supply chain of Fabletics, discussed how the omnichannel activewear fashion brand is leveraging RFID, AI and mobile technology to gain accurate intelligence about the location, status and history of products in its supply chain, with a particular eye toward meeting the wants and needs of Gen Z shoppers.
"Gen Z wants responsible products and fast delivery from a company that stands for something," said Portis. "They want speed, transparency and a verified, responsible supply chain."
According to Portis, Fabletics has instituted an RFID program in 100 stores that leverages integrated analytics made accessible by handheld RFID readers to let store associates see precisely where inventory is located.
"Employees know when a product will be available and where it is located across the supply chain," Portis stated. “RFID tags give each product a unique digital ID.”
Fabletics also integrates RFID tag data with AI-based predictive analytics to sharpen its last-mile delivery efforts.
"Last-mile delivery is not transactional anymore," said Portis. "Customers expect same-day delivery and services like lockers and curbside pickup. With predictive models, we can position inventory near stores where the demand is, and associates can check for the availability of out-of-stock products for same-day delivery or pickup at a locker or curbside."
In addition, by enabling individual products to be tracked back to the sources of their raw materials, Portis said RFID meets Gen Z desire to purchase products they know are sustainable.
Other benefits include having an "early warning" system to provide alerts of product issues such as substandard sewing or overproduction, said Portis.
Nestle – Middle mile automation
Jesse Chen, director of logistics for Nestle, reviewed how the CPG giant focuses on the middle mile (where goods travel between warehouses or from a warehouse to a last mile pickup point) of its direct-to-consumer supply chain to consolidate its distribution centers.
"Autonomous trucks transform the middle mile into a conveyor belt with faster, cheaper delivery," said Chen. "It used to be time- and resource-intensive. But autonomous trucks run non-stop with a shorter travel time."
[READ MORE: Taking charge of the supply chain’s last (and middle) mile]
As a result, Chen said Nestle has been able to make direct-to-consumer deliveries of goods such as HelloFresh meal kits and Zazzle custom products within one day 80% of the time. Utilizing automation of the middle mile and within distribution centers, the company plans to eventually serve the entire country with one-to-two-day delivery of direct-to-consumer orders from between one and three facilities.
EasyPost – How to win with AI implementations
Lori Boyer, head of content for parcel shipping software provider EasyPost, outlined strategies retailers can use to achieve victory with their AI technology rollouts.
"Retailers are good at vertical AI deployments that address a real pain point," said Boyer. "Not wide deployments. Too often they treat AI as a shiny object. Just because an AI solution has a good demo doesn’t mean it will help you."
Other AI pitfalls Boyer warned retailers to avoid include the “data perfection trap,” or not allowing AI solutions to gradually learn from the data input into them; and building your own AI tools rather than buying third-party technology.
"Start now, win fast," advised Boyer. "Buy, don’t build, AI solutions and layer them into existing systems. Focus on ROI, not hype."
This means focusing logistics AI projects on strategic wins that can be reached within roughly 90 days, said Boyer, giving examples such as route optimization and invoice auditing.
"Make the pain point as narrow as possible," Boyer said. She also urged retailers to welcome AI projects that disrupt current processes.
"It’s a red flag if AI implementation is too easy," she said. "It’s not supposed to be comfortable at first."
