Why real-time temperature monitoring, and when it isn't worth it

The benefits: what changes when you find out in transit

An excursion can still be cut short

A device read on arrival tells the receiving site what happened after it is over. A device that reports in transit raises the alarm while the excursion is still happening, when a pallet can still be moved off an apron, a container's power restored, or a delayed shipment put on another flight. The World Health Organization (WHO) already asks for arriving shipments to be moved off the airport apron or the wharf into temperature-controlled storage as soon as possible; a device that reports from the apron tells someone when that has not happened.

The Medicines and Healthcare products Regulatory Agency (MHRA), the UK regulator, puts the case in risk terms: making an excursion more likely to be detected gives earlier warning, may help prevent it, and gives greater assurance that conditions were maintained. Why the apron, customs, and handovers are where excursions cluster is in our guide to temperature excursions.

Distica, which distributes medicines in Iceland, has put it more simply:

The record is complete when the shipment lands

Product that arrives does not go into saleable stock until its transport record has been reviewed and the product released. With a device read on arrival, that review waits until someone has found the device, read it out, and filed the result. If the read-out shows an excursion, the product stays in quarantine until it has been assessed.

When the record is already in the system as the shipment lands, the review can start at once, and where the rules allow it, it can be automatic: a product that stayed within its stability budget, including through an excursion the budget absorbs, needs no one to decide. At Roche, that approach has automated up to 99.5% of product releases with no human intervention, according to a joint presentation at LogiPharma Europe in 2023. How it works is on our page on Zero-Touch Release, and what any automated release needs in place, whoever provides it, is in our guide to product release after transport.

You know where the shipment is

A device that reports its location answers the other question a delay raises. World Courier has told the story of a medicine shipped from Europe to a single patient in São Paulo, which an airline flagged at a transit airport because of a mismatch in the cargo records. The shipment's device showed exactly where it was, World Courier gave that to the airline, and the discrepancy was resolved within 20 minutes. The shipment continued to São Paulo with minimal delay, and the patient's treatment stayed on schedule.

A record with a location also says where an excursion happened: which airport, which handover, which leg. That is what an investigation needs first, and it is what turns a single excursion into a fix for the next shipment on the same lane. The case for keeping a lane's risk current from every shipment that runs on it is in Why lane risk assessment is the cleanest first use case for AI in pharma.

Many shipments, one decision

When something goes wrong across a network rather than on one shipment, real-time data tells you which shipments are exposed while there is still time to act. In a webinar with Controlant in May 2026, MSD's Ruud van der Geer recalled a Saturday morning in early 2024, when several dozen MSD shipments were in transit through corridors affected by the outbreak of conflict in the Middle East. The team had less than 30 minutes to assess their exposure and decide which needed immediate intervention. Every shipment was rerouted, rebooked, or held in safe intermediate storage.

The platform gave the team the picture. Deciding which shipments mattered most was still a judgment made by people:

The limits of real-time monitoring

Reporting in transit is not a guarantee. Five limits are worth knowing before you choose it.

It prevents nothing on its own

an alert helps only if someone sees it, at any hour, and is allowed to act on it. As the webinar with MSD put it, without clear ownership of those decisions, real-time alerts become noise rather than signal.

Alerts have to be set well

the MHRA has described a case where poor alarm settings meant no alarm was triggered, which gave false assurance that stock had stayed within its range. More alerts are not better ones.

Real-time has gaps

in the air, many devices switch their transmitter off to meet aviation rules, and on some ocean and remote legs there is no network. A device keeps recording and sends its readings when it reconnects, so what reaches the team is close to real time, not continuous, and on some legs nothing can be done until the next handover anyway.

It sees no more of the load

a real-time device takes a reading at intervals from one place in the load, like any other device. What changes is when the reading arrives, not how much of the shipment it measures.

It is not required

no regulation requires real-time monitoring. The EU Good Distribution Practice (GDP) guidelines and WHO guidance require that conditions in transit are monitored, recorded, and can be demonstrated, and leave the technology to the shipper.

What the rules do require in transit is in our guide to good distribution practice.

When a device read on arrival is the better choice

A device read on arrival records the same history for less, and for some lanes nothing more is needed:

• A robust product on a short, well-understood lane, where an excursion in transit would change nothing and the record only has to exist for the file.
• A product whose value is small next to the cost of acting on an alarm.
• A lane where no one could intervene in time even if they knew, and where release does not wait on the record.

The question is not whether real-time is better in general, but whether, on a given lane, knowing sooner would change what happens to the product or how fast it is released.

Questions to ask before choosing real-time

1. 1

   Who watches, and when?

   Alerts arrive at night and at weekends. Decide whether your own team covers those hours or someone does it for you, and what happens when no one answers.

2. 2

   What are they allowed to do?

   A forwarder can move a pallet; only the quality function can decide what happens to the product. Write down who may act on which alarm, and how.

3. 3

   How are alarms set?

   Set limits and delays from the product's stability data and the lane, so that an alert means something. Review them as the lane's history builds.

4. 4

   Can the record support release?

   If release is to start on arrival, the system that holds the record has to be validated for that use, with an audit trail.

5. 5

   Will the device fly, and come back?

   Battery-powered devices need the airline's approval, and reusable devices need a way home.

The rules for devices in air cargo are in our guide to pharmaceutical transport.

Where Controlant fits

Controlant's Saga devices report over the cellular network while the shipment moves, and our MARS team watches shipments 24/7/365 against the limits a customer's quality team sets. When an excursion alarm fires, MARS reaches out to the forwarder and escalates to the customer's contact, by the customer's work instructions. The record is on Controlant's GxP-validated platform when the shipment lands, so Zero-Touch Release can release product within its stability budget without waiting for anyone. The devices come back and are reused.

How this compares with the other approaches is on our page comparing cold chain monitoring approaches, and what each kind of device records is in our guide to cold chain monitoring devices.