© Lihin mohd |
Artificial intelligence could save around eight hours per air cargo shipment, by allowing truck dispatch and collection planning to begin before the traditional release-order trigger.
At Aviation Connect in Athens last week, CHI Cargo Group CEO Kai Domscheit said the air cargo industry was generating vast quantities of operational data, much of which remained trapped in emails, PDFs, scans, spreadsheets, and other unstructured formats.
“The future of air cargo is not going to be a single platform,” he said, arguing that the SaaS model was “dead”.
“It’ll actually be an ecosystem, with AI bots speaking to each other.”
But Mr Domscheit urged that the first step towards meaningful automation was not deploying AI, but making the underlying data usable.
“You need to get unstructured data structured. You need to put a context layer over it. Then you can start predicting. And once you’ve done that, only then you can start playing around with automation.”
The scale of the problem was illustrated by a single ULD, which Mr Domscheit said could generate more than 100 information flows at different stages of its journey, involving airlines, handlers, forwarders, airports, and other parties.
He highlighted several examples at Frankfurt, including release orders that indicated cargo was ready when operational checks had not been completed; unclear information about whether cargo was loose or in a ULD; and multi-master shipments where not all the necessary release forms were available.
The resulting inefficiencies included additional storage charges, empty truck runs, and wasted time, he said.
Mr Domscheit said handler CHI had analysed the sequence of events around import cargo at Frankfurt and found information was available well before the release order.
Under the current process, the release order acted as the trigger for dispatch, but he said AI could allow operators to anticipate the shipment, aircraft arrival, unloading and breakdown stages, and begin planning in parallel.
“We could take this and move it [up to] 10 hours,” he said. “We can work in parallel.”
However, despite these promising advances, Mr Domscheit cautioned against viewing AI as a straightforward replacement for people.
“We still have the human in the loop,” he said, adding that employees who embrace AI could become substantially more productive.
The presentation also struck a more cautious note on physical automation. Humanoid robots may eventually play a role in air cargo, but Mr Domscheit said the industry should focus first on preparing its operations and data.
“Treat data as a strategic asset, and prepare them properly – because it’s all about data.”