Temperature excursion management, explained.

Why 1,920 syringes wait on a truck dock until someone checks the batch's whole temperature history

From the alarm to stock a warehouse may pick again

1. The logger alarmsA reading crosses the product's limit, or a receiving check finds one.
2. The stock is heldThe batch, or the affected units, go to quarantine in the warehouse system.
3. The evidence is gatheredLogger files, the shipment, the batch's earlier time out of range, the product's stability data, the shipper's qualification.
4. A person decidesRelease, keep in quarantine or reject, signed, with the reasons written down.
5. The record and the fixStock status changes, the running total is updated, the carrier or the site is asked to stop it happening again.

Steps 3 and 4 are the subject of this guide.

A clock that never resets: four facts about excursions that newcomers do not expect

Excursion, allowance, running total, disposition

Fourteen terms are defined in the one-page summary.

Heat, freezing, gaps and returns: the six kinds of temperature excursion and the evidence each one needs

In shortAn alarm and an excursion are different things. An alarm says a sensor crossed a limit. An excursion is the product itself outside its approved conditions, and the evidence needed to judge it depends on what kind it was: a short warm spell, a long or repeated one, a freeze, a gap in the record, a return from a customer, or an alarm from an air probe while the product stayed in range.

Loggers can carry two kinds of probe. A product probe sits among or inside the cartons and stands for the medicine's own temperature. An ambient probe measures the air around the pallet, which swings much faster. An ambient alarm with every product probe in range is usually closed with a note, under the site's own procedure. The first question in any assessment is therefore which reading crossed which limit, for how long, and whether the record is complete.

EU good distribution practice (GDP) asks for "a procedure … for investigating and handling temperature excursions", and that excursions during transport are "reported to the distributor and recipient of the affected medicinal products".1 The World Health Organization's model guidance, written for time- and temperature-sensitive products, says such products should be quarantined if they "experienced unacceptable temperature excursions during transport" and not moved to saleable stock "until all relevant disposition procedures have been completed".4

The words people search for, and what each means

EU GDP, ICH stability guidelines, USP chapters and 21 CFR on excursions, with their status on November 3, 2026

In shortThe core texts are old and stable: the EU GDP guidelines of 2013, ICH Q1A(R2) of 2003 and Q5C of 1995, and the US current good manufacturing practice (CGMP) rules in 21 CFR Part 211. What is moving is around them: a new consolidated ICH Q1 that adds a section on excursions and is still a draft, a USP proposal on room temperature, an FDA proposal on distributor licensing that has not become final, and draft EU rules on computerized systems and AI.

USP chapters are sold by subscription. We read USP's public preview of the 2026 <659> proposal and the ECA Academy's reports on <1079.2>; we did not read the official chapters. Check the chapters' current text in USP–NF before quoting them in a procedure.

What a product's own stability studies allow after an excursion, and why two refrigerated medicines can have opposite answers

In shortAn excursion can be accepted only where the product's own stability data, or a scientific justification built on it, shows quality is unaffected. That data is specific: a temperature ceiling, a duration, whether time is counted once or in total, whether the product may go back into the refrigerator, and whether freezing is ever allowed. A general rule of thumb never replaces it.

Stability studies are the tests a company runs to set a shelf life and storage conditions. Samples are kept for months or years at the label condition, and at warmer accelerated and intermediate conditions, and tested for the attributes that matter: potency, purity, particles, appearance and others.

ICH Q1A(R2) says data from the accelerated and intermediate conditions "can be used to evaluate the effect of short term excursions outside the label storage conditions (such as might occur during shipping)".6 ICH Q5C adds that stress studies "may be useful in determining whether accidental exposures to conditions other than those proposed (e.g., during transportation) are deleterious to the product".2

The draft consolidated ICH Q1 puts this in one place. Its section 14.1 says: "Transient temperature excursions outside of the label storage conditions, may be acceptable if justified and supported by stability data." It continues that judging excursions along the supply chain "requires a comprehensive knowledge of the supply chain and an understanding of a drug substance and drug product's stability profile", and that "each excursion should be documented and handled within the corresponding quality management system or appropriate risk assessment".5

In the US, 21 CFR 211.166 requires that stability results "shall be used in determining appropriate storage conditions".17

The practical result is a short statement per product, approved by quality and kept in a controlled document: what range is allowed, for how long, how the time is counted, and what is never allowed. Some of it appears on the label, where patients and pharmacists can use it. Much of it stays in the company's stability summary for its own distribution decisions.

The four labels read differently on more than duration. One counts the "combined time out of the refrigerator". One gives a total "regardless of whether the product is in-use (opened) or before". One allows storage "one time" at up to 30 °C and says "do not return to the refrigerator". One sets a maximum period at up to 25 °C.25 These are patient-facing rules. A manufacturer's internal allowance for distribution may be narrower or wider, and only its own approved stability documents say which.

What a product's excursion statement usually holds

What a temperature data logger records, how minutes above the limit are counted, and where logger files mislead

In shortA logger file is a list of time-stamped readings from one probe at one position. Reading it well means knowing the probe type and position, the sampling interval, the time zone, the calibration status, the start and stop times, and whether any readings are missing. Minutes out of range come from the readings, under a counting rule your procedure sets.

In the fictional example from the top of the page, two loggers rode in the pallet: one in the top layer and one in the bottom, both product probes, each reading every 5 minutes for 391 readings with no gaps. The top-layer logger crossed 8 °C at 1:53 am and came back below it at 4:39 am, which is 2 hours 46 minutes. Its highest reading was 12.6 °C, at 3:10 am. The bottom-layer logger peaked at 11.1 °C. The lowest reading on any product probe was 4.2 °C, so freezing is ruled out.

Five checks before any number from a logger is used

Counting minutes out of range sounds mechanical, and still needs a rule. With 5-minute readings you can count the intervals whose reading is above the limit, or take the time from the first reading above to the first reading back in range, or interpolate between readings. The answers differ by a few minutes per crossing, and over a batch's life those minutes add up. Choose one method, write it into your procedure, and never fill a gap with a guess: a gap is unknown time, handled as in section 8.

Adding up a batch's time out of range across every leg, and working out a mean kinetic temperature

In shortFor a product whose allowance is a total time, the sum is simple and the inputs are scattered: every period the batch spent out of range in packing, storage, earlier shipments and returns, added to this trip and compared with the allowance. For room-temperature products, mean kinetic temperature (MKT) expresses heat stress over a period as one number, and it has limits that matter.

The running total in the example, step by step

The sample stability summary allows 70 hours in total between 8.1 and 23 °C over the shelf life, never above 23 °C and never below 0 °C. Two conditions must hold before the total means anything. Every reading must sit inside the allowed band: a peak of 24 °C would fall outside it even with hours left. And every period of the batch's life must have records: one period with no records makes the total unknown.

A batch with 7 hours left is still within its allowance, and a single long delay on its next trip would take it over. Teams that track the running total can route such batches to short, well-controlled lanes, or ask the stability scientist before they ship again. Teams that look only at each trip's logger cannot see the problem coming.

Mean kinetic temperature, with a worked example

ICH Q1A(R2) defines MKT as "a single derived temperature that, if maintained over a defined period of time, affords the same thermal challenge to a drug substance or drug product as would be experienced over a range of both higher and lower temperatures". It is higher than the simple average and "takes into account the Arrhenius equation", the relation between temperature and reaction rate, using the formula of J. D. Haynes from 1971.6

MKT is a tool with known limits. The ECA Academy, reporting USP's revision of chapter <1079.2>, quotes it: "MKT should not be abused, as MKT alone is not enough to assess the impact of a temperature excursion." The time, the actual temperatures, any exposure above 40 °C and the period used for the calculation must also be known.22

USP's commentary on the 2025 revision says "all available data must be used".23 Health Canada says MKT "may not be appropriate" where liquids or suspensions may change phase, where products need refrigeration or freezing, for biologics, and where data show excursions affect quality.3 USP's 2026 briefing on its <659> proposal adds that <1079.2> clarifies that storage time in colder conditions "cannot be used to offset excursions outside CRT ranges".7

For a refrigerated antibody, then, MKT rarely decides anything. The time and peak against the product's own allowance, and the freeze rule, do. For tablets labeled for controlled room temperature, MKT over a defined window is a common part of the assessment, and the product's own data still has the last word for any reading outside the label's range.

Was the shipper qualified for what happened? Reading a packaging qualification against a door-open, a delay or a hub transfer

In shortPassive shippers, active containers and refrigerated trucks are qualified by testing them against defined routes, seasons and events. An excursion inside what the qualification tested usually points to a known, limited risk; one outside it points to a gap in the lane, the packaging or the procedure, and often to a deviation.

EU GMP Annex 15 asks that "transportation routes should be clearly defined", that "seasonal and other variations should also be considered", and that a risk assessment covers "delays during transportation, failure of monitoring devices, topping up liquid nitrogen, product susceptibility and any other relevant factors".8

EU GDP says the choice of container depends on "the anticipated external temperature extremes", "the estimated maximum time for transportation including transit storage at customs" and "the qualification status of the packaging".1 It adds rules on cool-packs: no direct contact with the product, staff trained on "seasonal configurations", and a system "to ensure that incompletely cooled packs are not used in error".1

In the example, the carrier reported that the trailer doors were open for 52 minutes at a cross-dock in Bordeaux. The shipper's qualification report includes a 60-minute door-open challenge, in which the top-layer probe reached 12.9 °C and was back below 8 °C within 3 hours 10 minutes. The real event was shorter and milder than the test, which supports the release. It still earns a notice to the carrier, because a qualified shipper is a safety margin, and the door-open used part of it.

Cool-packs, hubs and seasons: where lanes fail

Three causes recur in the texts.

Frozen gel packs put straight into a winter shipper can drop the top layer below 0 °C within hours, which is why EU GDP asks for control of cool-pack conditioning and segregation of frozen from chilled packs.1 Transfers at hubs, where pallets wait on docks or loggers are removed and re-attached, are named in EU GDP: "particular attention should be paid to temperature monitoring" where a route includes "unloading and reloading or transit storage at a transportation hub".1 And summer and winter differ enough that Annex 15 and GDP both ask for seasonal profiles.8,1

Three cases where the allowance arithmetic stops: a freeze, a gap in the record and a return without storage evidence

In shortA freeze of a product that must not freeze is decided by the freeze rule, whatever time is left. A gap in the record means the total cannot be known until another record covers it. A return of refrigerated stock needs documented evidence for the whole time it was away, under EU GDP 6.3 and, in the US, 21 CFR 211.204 and 205.50.

A freeze in a product that must never freeze

When a product's stability data says it must not be frozen, any reading below 0 °C on a product probe is usually treated as a reject, and the allowance for heat is irrelevant. The useful work is finding why. In the fictional example's second case, the packing record shows gel packs taken from a freezer at 1:42 pm and the shipper closed at 1:51 pm, where the work instruction asks for two hours of conditioning. That is a site failure and gets a deviation with a corrective and preventive action (CAPA).

A gap in the logger record

A logger removed during a container swap, a battery that died or a monitor that went offline leaves time with no readings. Most procedures keep the stock in quarantine until a second record covers the gap: another logger in the same load, or an active container's own air log. Annex 15 names "failure of monitoring devices" as a risk to plan for.8 If no second record exists, the excursion is unknown in size, and the running total cannot be calculated.

A refrigerated return without storage evidence

EU GDP 6.3 lets products that have left the distributor go back to saleable stock only if all of five conditions are met, including unopened packaging, an acceptable time limit ("for example 10 days") for returns from customers without a wholesale authorization, and examination "by a sufficiently trained and competent person".

For products with special temperature conditions, "returns to saleable stock can only be made if there is documented evidence that the product has been stored under the authorised storage conditions throughout the entire time". The evidence covers delivery, examination, opening of the transport packaging, return to the packaging, collection, and return to the distribution site refrigerator. "If any deviation has occurred, a risk assessment has to be performed." The Responsible Person decides the final disposition of returned products.1

US rules say the same thing in their own way. 21 CFR 211.204 requires returned drug products to be held, and destroyed if their storage or shipping "casts doubt on the safety, identity, strength, quality or purity", unless "examination, testing, or other investigations prove" they meet standards.18 21 CFR 205.50 applies the same test to wholesale distributors.20 And 21 CFR 211.208 forbids salvaging products exposed to "extremes in temperature" from "natural disasters, fires, accidents, or equipment failures" without laboratory evidence and evidence from the premises.19

What FDA's citations, MHRA's GDP findings and the US recall record show about storage and excursion failures

In shortInspectors cite weak stability programs and storage conditions every year. UK distribution inspectors have found excursions recorded and never followed up, uncalibrated transport monitors and invented readings. Recalls that name temperature exposure are a small share of the US drug recall record, and they count recalls: an excursion decided inside the company leaves no public record, so no public source counts excursions.

The 148 is our sum: 120 citations across ten lines of 21 CFR 211.166 and 28 across two lines of 211.142. FDA's file lists 713 drug Form 483s issued from its system in that year, and says some 483s prepared by hand are not in it.26 One citation counts one observation, and an inspector's wording behind each one is not in the file.

What UK distribution inspectors wrote down

MHRA, the UK regulator, published its GDP deficiency data for 2016 from a sample of 148 of the 1,428 GDP inspections it carried out that year. Its examples include "Cold chain temperature deviations had been recorded but not subjected to CAPA reporting"; a refrigerator with a build-up of ice reading "0.6°C … at the time of the inspection"; "Uncalibrated equipment was being utilised to monitor temperature during transport"; and a site where "There were considerable gaps in contemporaneous temperature records. The Responsible Person confirmed he had added fabricated readings to fill gaps."27

We found no later MHRA GDP deficiency report. The examples are old, and each one is still a question to ask of your own records.

How we counted: FDA's openFDA recall data for drugs, events with report dates from January 1, 2020 to December 31, 2025, searched for reasons that mention temperature, refrigeration, freezing, excursions, storage or heat. Each of the 45 matches was read by hand, and 12 were left out because the reason was something else, such as a failed stability time point, a desiccant or a missing label text.

The 33 that remain cite exposure to temperatures or storage outside the label, in storage or transit.28 This is our count, and it counts recalls: it says nothing about how many excursions happen, since nearly all are decided inside the company.

The inspector general's 2012 study looked at doctors' offices and clinics, at the end of the chain, and it measured a cumulative total over two weeks.29 It is an old study of one program. Its lesson travels: when nobody adds up the hours, many small excursions go unnoticed, and the total can be far larger than any single event.

What an excursion puts at stake for a distributor or manufacturer: held stock, wasted batches, replacements and the record an inspector reads

In shortA wrong release puts out a batch whose quality its own data no longer supports. A wrong rejection destroys good stock and triggers a replacement shipment. A slow decision ties up stock and keeps a customer waiting. And a weak record turns a sound decision into an inspection finding. Temperature-controlled products are a large and growing share of medicines.

IQVIA's white paper on the vaccine cold chain says the global cold chain market "accounted for 38% market share of all pharmaceuticals, up from 26% in 2017, and expenditure on cold chain logistics was estimated to be around $21.3 billion in 2024".30 A figure of about $35 billion a year lost to temperature-control failures is widely repeated, often credited to IQVIA's institute and a 2019 survey; we could not read its original, and a September 2026 trade article repeats it without naming its source.31 Treat it as an industry estimate.

That survey counts respondents. It shows that excursions are common and gives no rate per shipment.32 No public source we found gives the share of excursions that end in release or rejection, the hours an assessment takes, or the cost per assessment. Those are numbers to measure in your own records, as section 18 shows.

Four ways an excursion decision goes wrong, and what each costs

For licensed biologics in the US, FDA expects a manufacturer to report a biological product deviation when an event in "storage, or with the holding or distribution" of a distributed product departs from requirements or specifications in a way that "may affect the safety, purity, or potency" of the product, under 21 CFR 600.14.33 That is a quality report about the product, and it is one more reason the record behind each excursion decision has to be complete.

What software and AI can do in an excursion assessment, what EU draft Annex 22 and FDA say about it, and what stays with quality

In shortThe slow parts of the work are mechanical: reading files, matching serial numbers to batches, adding up minutes across systems, finding the right page of a qualification report, drafting a notice. Software can do those well. The decision, any judgment the product's data does not settle, and any change to an allowance stay with people, and the EU's draft rules on AI in GMP make that boundary sharper.

The draft calls an application non-critical where it does "not have direct impact on patient safety, product quality or data integrity".24 Whether an excursion assessment is a "critical GMP application" under draft Annex 22 is a question each company has to answer and write down, with its quality unit, while the text is a draft.

A design in which software reads, adds up, cites and drafts, and a named person checks the evidence and signs every disposition, keeps the decision with people. Whatever the tool, its output becomes evidence a quality decision relies on, so it sits inside your validation and your data integrity controls, with an audit trail of who did what.

Who handles a temperature alarm today, how the files move between them, and where the hours of an excursion assessment go

In shortOne alarm involves the GDP specialist, the person who signs, a stability scientist, logistics, the carrier and often the customer. The files move by email and shared drives. Most of the time goes on finding records and adding up earlier exposure; the decision itself is usually quick once the evidence is in one place.

The people one alarm reaches

Three ways teams organize the work

Alarm EX-2614: Veltrimab syringes at 12.6 °C on the road to Madrid, from 2:24 am to a signed release at 8:14 am

In shortThis is the example from the top of the page, inside the product. Specialist agents, AI helpers that each do one job, read both logger files, tie them to the batch, add the earlier legs to this trip, check the shipper's qualification and draft a cited recommendation. Dana Okafor, the GDP quality lead for Europe, reads it and signs.

1

The new alarm arrives on a wall of open alarms

Dana's alarm wall shows the logger feed live with 412 shipments in transit, six open alarms and €9.55 million of stock on hold. Each alarm is one card: the trace against the label band, the stability budget, the recommendation and a 24-hour clock, which is the sample company's own procedure. EX-2614 is new: Veltrimab PFS (prefilled syringes), 1,920 syringes, peak 12.6 °C, budget not yet counted, allowance 70 h at up to 23 °C.

Beside the cards, the batches closest to the end of their allowance are listed under the line "time out of range is cumulative and never resets".

2

One press, and the agents work on the card itself

Dana presses Assess. The logger reader takes both files, 391 readings each: peak 12.6 °C, 2 h 46 min out of range. The shipment matcher ties them to SH-77502, batch VL26-0388, 1,920 syringes, Quorlan Freight. The stability budget keeper reads the allowance and adds the earlier legs.

Each step shows on the card as it runs, with what it found.

3

95% of the budget left, and a recommendation to release

The budget meter fills: 3 h 21 min of 70 h used, 66 h 39 min left, 95%. The reason reads "Highest reading 12.6 °C for 2 h 46 min above 8 °C, never below 2 °C". The card moves to Ready for signature with "Release · 91%" and 22 h 30 min left on the clock.

The confidence figure is the drafting agent's own estimate of how firmly the evidence supports its recommendation. It is no substitute for reading the evidence.

4

The trace, both loggers, and every leg of the batch's life

Dana opens the alarm. The temperature record plots the top-layer logger LG-4471 peaking at 12.6 °C and the bottom-layer LG-4472 against the 2 to 8 °C band. Below it, the batch's time out of range leg by leg: 35 min in the packing hall at Plant 3 on Sep 30, 0 min in the cold room from Sep 30 to Oct 5, 2 h 46 min on this shipment.

On the right, the logger rows sit open: every reading from 12:40 am, the peak highlighted.

5

Four checks, each with its source

Four checks sit under the legs: never below 2 °C, no freezing (lowest 4.2 °C, on the bottom layer); inside the stability allowance (peak under the 23 °C ceiling, 3 h 21 min of 70 h used); shipper behaved as qualified (52-minute door-open against a 60-minute test that reached 12.9 °C); complete temperature record (two loggers, 391 readings each, no gaps).

Each small number is a citation. A click opens the source at the passage.

6

The recommendation, with its reasons and what follows

"Release to saleable stock", with four numbered reasons: the peak and time; the allowance ("Veltrimab may spend up to 70 h in total at up to 23 °C over its shelf life"); the running total, "including 35 min at Plant 3 on Sep 30", which "is cumulative and never resets"; and the door-open inside the qualified challenge. Then the conditions: write 3 h 21 min to the batch log, send the carrier a deviation notice, set the ERP batch status to unrestricted.

Below it, the agents' own trace: what each read, in what order, how long it took.

7

The door-open, beside the qualification's own words

Dana clicks the citation for the door-open reason. The packaging qualification PQ-0217 for the PalletCool 1200 opens at the passage: "During the 60-minute door-open challenge at hour 24 the top-layer probe reached 12.9 °C and was back below 8 °C within 3 h 10 min; all other positions stayed within 2–8 °C."

The panel's note says the sources are read by the agents and never changed.

8

A carrier deviation notice, drafted from the evidence

Because the cause was the carrier's, the deviation drafter writes notice CN-26-0412 to Quorlan Freight: what the loggers recorded, from 1:53 am to 4:39 am on Oct 7, while the doors were open at its Bordeaux cross-dock, and a request under section 7.3 of the quality agreement for the root cause and corrective action by Oct 14, 2026. Dana can edit it before it goes.

9

She signs, and the signature states what it means

Dana chooses Release, as recommended. The dialog states the meaning, "I release this stock to saleable stock", for Veltrimab 150 mg/mL prefilled syringe, batch VL26-0388, 1,920 syringes, and asks for her password. It notes that the signature's meaning, name, date and time are stored with the record, which is what 21 CFR 11.50 asks a signed electronic record to show.38 Choosing differently from the recommendation would ask her to say why.

10

Back in saleable stock, the running total written, logistics told

The alarm closes: "Released to saleable stock · e-signed by Dana Okafor · Oct 7 8:14 am". The ERP batch status is set to unrestricted, 3 h 21 min goes on the batch log for VL26-0388, and Sam Patel in the logistics control tower hears that no replacement is needed.

A frozen autoinjector, a return with 39 hours unrecorded and a logger removed at a hub: three alarms the allowance cannot settle

In shortThree other alarms on the same wall show where the arithmetic stops, as described in section 8. A freeze is a reject whatever time is left, with a deviation and a second signature. A return without fridge records cannot be counted, so it cannot go back to stock. A gap in one logger keeps the stock in quarantine until another record covers it.

EX-2611: −1.8 °C, and the freeze rule decides

Veltrimab 100 mg autoinjectors, batch VA26-0142, 2,400 pens, €3.6 million, from the EU distribution center in Venlo to Nordvik Wholesale in Oslo with Quellmoor Logistics. The top-layer logger read −1.8 °C and stayed below 0 °C for 65 minutes, in the first two hours after packing. The screen says it plainly: "The freeze rule applies. Veltrimab must not be frozen — the time budget does not matter for this decision."

The recommendation is "Reject — freeze event" at 97%, with four cited reasons: the reading, the product's freeze rule and the procedure that applies it, the packing record (gel packs out of freezer F2 at 1:42 pm, shipper closed at 1:51 pm, against two hours of conditioning in work instruction DC-114), and the winter qualification's own result for packs used unconditioned. A reject needs a second signature from Omar Haddad, the Responsible Person for the EU wholesale license.

The deviation drafter writes DEV-26-1179, class major: what happened, why, the impact on 2,400 pens, a check of two other winter shipments packed by the same shift, and a proposed CAPA of a conditioning timer, a pack-out sign-off and refresher training by Oct 21, 2026. A person files it.

EX-2608: a return that cannot be counted

Nordvik Wholesale returned 200 Veltrimab syringes of batch VL26-0371, €300,000, as an over-order ten days after delivery. Its fridge monitor was offline from Oct 2 at 6:00 pm to Oct 4 at 9:00 am, 39 hours, and no manual readings were taken. The budget line reads "No total — 39 h without storage records": the running total cannot be calculated.

The recommendation is to reject at 95%: refrigerated returns go back to saleable stock only with documented evidence for the entire time, under EU GDP 6.3, and the Responsible Person decides on returns.1 Dana has signed; the card waits for Omar Haddad's second signature.

EX-2612: a logger removed at a hub

6,000 vials of Dravenzo 20 mg, batch CR26-1098, €2.1 million, flew from Leiden to Lagoon Health in Singapore with Skerrow Air Cargo, via Dubai. During the container swap at the hub, the pallet 1 logger LG-5120 was removed and re-attached, leaving 3 h 40 min with no readings. Pallet 2 in the same container peaked at 9.8 °C, 1 h 50 min above 8 °C, well inside Dravenzo's 14-day allowance.

The sample procedure keeps stock in quarantine for gaps over 30 minutes until a second record covers them, so the recommendation is to keep it in quarantine at 88% and ask the carrier for the active container's own air log.

Six specialist agents read loggers, count every leg and draft; a GDP quality lead signs, and the Responsible Person co-signs rejects and returns

In shortEach agent does one job and has written limits. None of them signs, releases stock, changes an allowance or files anything unreviewed. People decide, under the sample company's procedure, and the four-eyes rule applies to people.

Dana, Omar, Lena, Priya and Sam in the screens

The excursion list, lanes and carriers, product allowances, the batch log, the dashboard and the rules

In shortAround the alarm wall sit the screens a quality lead uses every week: every alarm in one table, lanes with their excursion rates, each product's allowance with its stability source, every batch's running total, the trend dashboard, and the procedure's rules and signers as settings.

From logger files, shipment records and stability summaries to a signed disposition, and what the solution is designed to reach

In shortThe solution reads the records a cold chain already keeps and writes back only after a signature. Its outcome numbers below are design targets and estimates on sample data, to be measured on your own excursions.

The systems an excursion's evidence already lives in

Built for scale across sites, lanes, carriers and products, with roles, maker-checker approvals (one person prepares, another approves), a full audit trail, export and integrations configured for each network. Your own products, lanes, rules and people replace everything shown here.

What an excursion costs your network in stock on hold, replacement shipments, wasted batches and inspection exposure

In shortWe found no independent figure for the hours an excursion assessment takes or for the share of excursions released, so this section shows you how to find your own in a week. Start with five numbers from your own records: excursions a year, median hours from hold to signature, value on hold, replacements booked while waiting, and stock rejected.

Finding your own excursion baseline in five working days

Specialist fatigue, a customer's trust and the leg nobody added

What faster excursion decisions return: hours of quality time, days of stock on hold and replacement shipments not booked

In shortPut in your own numbers. The calculator values quality hours you will reuse, days taken off the hold, and replacements you no longer book, then subtracts the yearly cost. Every default is an assumption to replace, and the value rests on your own baseline.

Quality hours, hold days and replacement shipments: where the value comes from

Excursions a year, hours per assessment and value on hold

At these assumptions a site with 50 excursions a year does not cover the cost, and may be better served by a clean procedure, a single running total per batch and a weekly look at lanes. A network with hundreds of excursions clears it, and its first question should be what the released hours are for. A low-value product changes the picture too: the carrying cost line fades and the case rests on hours and replacements.

How ready your distribution quality desk is: excursion volume, approved allowances, running totals and a QA position on AI

In shortPick the answer closest to your network today. Eight questions score 0 to 2, and the last is for reflection. The meter adds up your score as you go, and each question says why it matters and what a strong answer sounds like.

What to ask a temperature excursion software vendor about allowances, running totals, signatures and draft Annex 22

In shortAsk every vendor the same questions, us included. The answers show whether a tool only flags logger breaches or assesses them against each product's own data, whose rules it applies, what it records, and how you would test it on your own excursions.

A ten-week excursion pilot: past alarms first, then one lane live, with your quality lead signing as usual

In shortTake 30 past excursions with known outcomes, including a freeze, a gap and a return. Write each product's allowance as an approved statement. Run one lane's live alarms in parallel, with your desk working as usual and your signers deciding every one. Decide on three numbers: totals missed, wrong citations and hours per case.

1. Measure first: time five assessments and pull alarm-to-signature hours for your last 50 excursions.
2. Write the allowances: one approved statement per product with your stability group: band, ceiling, total, freeze rule.
3. Replay 30 past excursions, including copies where you plant a missed earlier leg, a gap and a freeze.
4. Run one lane live, in parallel, for four weeks. Your desk works as usual and your signers decide from your current process.
5. Decide on your numbers, with a written position from QA on Annex 22, a sampling plan and your validation approach.

Totals missed, wrong citations and hours per excursion

Track all six measures. Totals missed, wrong citations and hours per case are the three you decide on.

Temperature excursion questions answered: allowances, EU GDP, mean kinetic temperature, returns and software

In shortQuestions distribution quality teams and buyers search for, each answered on its own.