Artificial intelligence could cut roughly eight hours from each air cargo shipment by letting truck dispatch and collection planning begin before the traditional release-order trigger fires.

That is the claim made by Kai Domscheit, CEO of CHI Cargo Group, at the Aviation Connect conference in Athens last week. His argument rests on a simple operational constraint: today, collection planning waits for a release order, and everything upstream of that moment is dead time AI can eliminate.

Domscheit's diagnosis of the industry's data problem is blunt. He told the Athens audience: "The future of air cargo is not going..." — his framing, as reported by The Loadstar, points to an industry generating vast quantities of operational data that remains locked away in unusable formats.

The bottleneck, in his telling, is not data scarcity but data accessibility. Air cargo operations produce enormous volumes of information, yet much of it sits trapped in:

• Emails
• PDFs
• Scans
• Spreadsheets
• Other unstructured formats

What does the eight-hour figure actually mean?

The claim is a cycle-time saving, not a transport-time saving. The mechanism is anticipatory planning: if AI systems can read unstructured inputs — shipment notifications, bookings, status emails — and infer when cargo will be ready, truck dispatch can be scheduled before the formal release order arrives.

For ground handlers and forwarders, the commercial stakes are familiar. Trucks dispatched earlier and more predictably compress the door-to-airport segment of the shipment timeline. An eight-hour reduction per shipment, if sustained across a handler's volume, translates into meaningful capacity gains without adding vehicles or drivers.

The figure is a projection from a handling executive, not a verified operational result. As with manufacturer performance claims, it should be weighed against deployment evidence — which handlers are running such systems in production, on what volumes, and with what measured cycle-time deltas. The Loadstar's report does not cite pilot data behind the number.

Why unstructured data is the target

The industry's dependence on email, PDFs and scans is a long-standing friction point. Machine-readable data standards exist across the cargo chain, but day-to-day operations still run through documents humans read and re-key. Domscheit's point is that AI has now reached the stage where it can extract and act on that trapped information rather than requiring the chain to be re-engineered first.

That reframes digitization. Instead of waiting for every participant to adopt structured formats, handlers can apply AI on top of existing document flows and start planning earlier — an incremental path that fits an industry with mixed IT maturity across forwarders, handlers and truckers.

The competitive question for handlers

For a group like CHI Cargo, the pitch doubles as positioning. Handlers that compress collection cycles can promise shippers faster, more reliable pickup-to-airport times, a differentiator in a market where airline capacity and trucking capacity are both tight at peak.

The unanswered question is adoption speed. AI-based planning tools require integration with customer booking flows, trucking fleets and warehouse systems — and the savings Domscheit describes depend on the release-order step being the binding constraint, which may vary by trade lane and customer.

Eight hours per shipment is a headline figure that handlers across the sector will now be pressed to test against their own operational data.