As AI transforms global logistics, David Weinberg, Head of Portfolio Management IoT Solutions, G+D Mobile Security explains why continuous, verifiable data is essential to improving supply chain visibility, strengthening accountability and enabling intelligent decision-making.
An AI-powered routing system issues a delivery optimization recommendation at 6 a.m. The location data it relies on is 14 hours old. The asset the system is optimizing for has already been sitting idle at the wrong terminal since midnight. No alert was triggered. No one knew.
During these trips, high-value products can be damaged, perishable goods can be exposed to extreme temperatures or contents can go missing. And yet, operations leadership may be none the wiser. At a time when cargo theft is also on the rise around the world, there is a clear need to close these visibility gaps and give enterprises greater oversight.
For AI-driven organizations, this is no longer only a visibility problem. It is an evidence problem. The question is no longer whether data exists somewhere in the supply chain, but whether that data is continuous, verifiable and trusted enough to drive automated decisions and establish accountability when something goes wrong.
Even though the global supply chain handles goods worth more than US$26 trillion annually, many organizations still rely on fragmented scan events and last-known-location data when making critical operational decisions.
Today, operations leadership often depends on logistics partners for information. Visibility programs typically integrate with partner systems to access this data. But these integrations do not remove the blind spots already inherent in transportation operations, as cargo passes through the supply chain, from haulage drivers to airport ground handlers, warehouse operators and beyond.
Closing those blind spots requires more than connecting existing systems. It requires the physical asset itself to become a trusted source of real-time information.
At its core, the system is still built around scan events that provide little more than a shipment’s last known location. That poses a challenge for organizations deploying AI-powered systems dependent on accurate data.
AI can only generate meaningful insights when supplied with information that is accurate, timely and verifiable. Without it, supply chain leaders lack the evidence needed to optimize routes, improve forecasting accuracy and drive efficiency across downstream operations.
The visibility gap is structural, not technical
What is required is always-on monitoring that does not merely estimate the location of goods in transit but can pinpoint it on a map.
Modern devices equipped with multiple layers of tracking technology, including GPS, cellular networks and Wi-Fi positioning, help provide this. They enable continuous monitoring of cargo and equipment, even in dense urban environments, warehouses and distribution centers where satellite signals may be weak.
By removing a single point of failure, these devices also help counter criminal tactics such as GPS disabling, which has been documented by TAPA and CargoNet as an increasingly prevalent tactic targeting intermodal assets.
In North America alone, cargo theft losses reached an estimated US$725 million in 2025, according to Verisk CargoNet analysis, with rail yards and intermodal terminals identified as primary target environments.
Advanced sensor technology also enables supply chain leaders to monitor the condition of goods while they are in transit. This is particularly relevant for pharmaceuticals or explosive materials, which are governed by strict regulations such as ADR, ATEX and FDA DSCSA.
While telemetry data is available across air, sea, rail and road transport, sophisticated sensors can also turn non-powered items, such as containers, pallets and even individual parcels, into connected assets. This makes it easier to track goods as they move from vehicle to vehicle.
These tracking devices are durable and long-lasting. Low-power connected devices, such as solar-powered trackers or credit-card-sized smart labels, have reduced the infrastructure dependency that historically made continuous monitoring impractical at scale.
Despite their size, they can provide extensive insight into how a package has been handled: whether it has been tampered with, exposed to temperature or light changes or used efficiently as part of the broader asset network.
When sensor data becomes accountability evidence
When this level of tracking data exists, supply chain leaders can make decisions in real time. If goods move when they should not, they can immediately investigate the incident and take corrective action.
If the doors of a reefer container are opened unexpectedly, an organization using AI-powered systems can identify the event and act before temperatures exceed permissible limits for too long.
This can have significant commercial implications. For example, Fiskskaldjur is a European company that specializes in transporting seafood delicacies. If its products do not remain within an optimal temperature range of 3–6 degrees Celsius throughout a journey, shipments can be spoiled, at a cost of EUR 100,000 per shipment.
Similarly, pharmaceutical products such as vaccines must be kept within a defined temperature range to remain safe and effective. A failure to identify temperature deviations in transit could ultimately lead to legal and regulatory action, along with significant reputational damage.
Constant monitoring means supply chain leaders are no longer reliant on manual checks by drivers. Companies automatically receive verifiable data that can help resolve disputes over liability after an incident.
If the contents of a parcel go missing during transit, the enterprise can identify where and when the package was tampered with and use that data to clarify responsibility.
Why poor tracking data actively degrades AI
As enterprises seek to optimize supply chain operations, improve forecasting accuracy and manage disruptions in real time, AI is playing an increasingly important role. But these capabilities depend on access to accurate data.
Models trained primarily on scan events and last-known-location data will have limited potential. Scan events provide snapshots of what happened in the past. AI-driven decision-making requires visibility into what is happening now.
Real-time location and condition data enables more detailed analysis, improves forecasting accuracy and helps organizations make decisions.
To realize the full potential of AI in logistics, cargo must become an always-on connected asset capable of supplying verifiable information. Only then can organizations reduce the blind spots that currently limit automated decision-making.
As supply chains become increasingly AI-driven, organizations can no longer rely on fragmented scan-event visibility alone. They need continuous access to trusted location and condition data that improves decision-making, strengthens resilience and enables more efficient operations.
Turning cargo into a connected asset is therefore not simply a visibility initiative. It is the data foundation that next-generation supply chains will depend on.
That also changes where responsibility sits. If AI systems are expected to optimize routes, predict disruption and support accountability, then the quality of the data feeding those systems becomes a strategic technology decision.
Deciding which assets generate data, and what quality that data must meet to be trusted, is therefore not only a logistics question – it is a CIO decision.
