AI Services for Logistics, Built on Real Operations

Intelligent Documentation & Customs

Document handling is the first place AI changes daily work, because the input is already text and the rules are already written down. Recognition plus rule checking removes re-keying and catches inconsistencies before submission.

• Invoice, packing list, bill of lading and declaration parsing into structured fields
• HS code candidates with reasoning, reviewed by a classification officer
• Cross-document consistency checks before the file is lodged

Our position: Document recognition and HS code assistance are in pilot, with officer confirmation required on every output.

Predictive ETA, Capacity & Rates

The next shift is from reporting what happened to anticipating what will happen: arrival times, equipment availability and rate movement, expressed as a range with a confidence level rather than a single promise.

• ETA ranges that update as milestone events arrive
• Equipment and slot availability outlook per lane for booking decisions
• Exception early warning when a leg drifts outside its expected window

Our position: Milestone data is being standardised first, because forecast quality depends on event consistency more than on model choice.

Visual Intelligence in Yard & Warehouse

Cameras add a layer that structured data cannot provide: what the cargo actually looked like at each handover, and whether the work around it was done safely. The value is evidence and prevention, not surveillance of individuals.

• Container and pallet condition, seal verification and loading checks
• PPE compliance, forklift and pedestrian conflicts, restricted zone intrusion
• Stacking height, lashing anomalies and yard vehicle speed

Our position: A limited-area pilot with officer review of every automated detection, signage at monitored areas and defined retention periods.

Agentic Logistics Operations

Agents move beyond answering questions to completing bounded tasks: collecting missing cargo details, preparing a quotation draft, chasing a milestone, or assembling a claim file. They work through defined interfaces and hand over to a person when judgement is required.

• Task-oriented agents for enquiry, documents, tracking and safety review
• Tool and system integration with permissions and a full audit trail
• Human-in-the-loop takeover, canary release and rollback at version level

Our position: Agent templates are being built on our own workflows first, then offered to industry customers through the AI cloud platform.

Route Optimisation & Emission Awareness

Cost and carbon are increasingly decided together. Mode selection, consolidation and routing choices all carry an emission consequence, and customers are starting to ask for the number alongside the rate.

• Mode and routing comparison including transit, cost and emission estimates
• Consolidation planning that reduces partially filled movements
• Per-shipment emission records for customer reporting requests

Our position: Emission estimates are being added to routing comparisons as indicative figures, clearly labelled until the calculation basis is agreed with customers.