When it comes to applying artificial intelligence to logistics and the supply chain, the headlines might be slightly ahead of real-world use.

A new survey by Freightos, the online freight booking and payments platform, finds that AI is still mostly a matter of talk among supply-chain executives today. Nevertheless, there’s significant interest in deploying the technology to make key decisions in the years ahead.

Freightos surveyed 55 respondents between July 26 and August 3. Half were importers and exporters, a quarter were freight forwarders and the rest were “other logistics professionals,” according to chief product officer Ruth Amaru.

Ninety-six percent of the relatively small sampling said they expect to adopt AI for some aspect of supply chain management. Yet just 14% were already using it, or piloting projects, for automating pricing or customer service. And only 7% were current “active” users of AI.

As for the expected impact of AI on logistics, more than half of the respondents said it will be “major”; 43% predicted “real, but limited impact,” and 5% called it “just hype.” That last group consisted entirely of small and medium-sized businesses, according to Freightos.

Amaru is pleased that only a small percentage of executives dismiss AI entirely, although she calls the small number of active users “a little surprising.”  

The logistics business “is a bit slower to adopt new technology,” she explains. “It’s a complex and very interdependent industry with many players.”

What might jumpstart AI projects in logistics is a parallel trend toward automating, or digitizing, key freight functions, especially pricing. Ten years ago, Amaru says, “people in the logistics industry didn’t understand what it meant to digitize.” More recently, the sector has experienced “huge progress” toward automating systems for displaying available freight capacity and booking online. As those applications become more widely embraced, the accompanying move to AI seems only natural.

Progress varies according to the mode of transportation. Amaru says the airline industry, which allows customers to search for capacity, view pricing and book online instantly through multi-partner reservation platforms such as Sabre and Amadeus, is well-advanced in reliance on automation.

Ocean carriers are less so. One reason is their smaller numbers — roughly 10 major carriers, versus hundreds of air-freight providers — with each of the players “feeling like they could win the battle.” That ambition has led to a proliferation of proprietary pricing and tracking systems, and a reluctance to share critical data across competitors.

Ocean carriers are now moving toward creation of an industry standard for data exchange. “But while they’re waiting for that to happen, each is working separately, and not willing to work [together] with technical partners,” Amaru says.

With the move toward digitization comes a keener interest in AI, which becomes necessary for managing the massive amounts of data that flow through modern-day supply chains. According to the Freightos survey, logistics professionals are most interested in applying AI to pricing, followed by shipping operations and customer service. Further down the list comes “e-mail insights,” software engineering, sales, and management of exceptions to regular service patterns.

Logistics is one of the last industries to apply AI to pricing, which has become increasingly complex as service providers seek better intelligence about their customers. “It’s very easy for everybody to understand that there’s no ‘correct’ price,” says Amaru. “There’s one that’s more effective, depending on what your targets are. It’s trial and error.”

A decade ago, the notion of putting international freight rates online was unthinkable, she adds. But the examples of other industries successfully taking that route are making the technology more attractive for all transportation modes.

The big question for logistics, as for any industry looking to adopt AI, is how it will affect the human worker. Surprisingly, the Freightos survey finds more than half of respondents predicting “minor” impact on logistics headcounts, while about a third saying it will be “major,” roughly 10% saying AI will actually create jobs, and around 5% foreseeing no impact at all. Larger companies expect a stronger impact on headcount than SMBs, Freightos said.

Despite its relatively slow acceptance by logistics and the supply chain, AI seems poised to play a key role in the industry over the coming years. But some refinements in the applications might be necessary first.

“We need logistics technology companies to produce things that are easy to use and work well,” Amaru says. “The adoption of each system makes it easier to adopt the next one.”