Reference

Technical notes

Five working notes published in 2026 by the Sglearninghubplus training desk. Each one is the sort of argument we rehearse on MAP-210, PSL-120, AIR-150 and EXC-130: how to read a file, where a calculation stops, and what still needs a named decision. Dates below are publication months, not intake dates.

March 2026

8 min · Sglearninghubplus training desk

Reading a mapping report: what the worst-case sensor actually tells you

A mapping report is often filed as proof that a cold room “passed”. That reading skips the only sensor that should change how the room is used. During empty-chamber and loaded mapping, loggers sit at geometric extremes and at suspected weak points: evaporator discharge, door reveal, highest pallet face, and the return-air path. The sensor that spends the longest time nearest the specified limit is the worst-case location. It defines the usable envelope, not the average of every channel.

If that logger sits at 7.6 °C in a 2–8 °C room for three hours after a defrost cycle, the report is not a near-miss on a summary page. Product stored at that coordinate will see a different kinetic load from product in the centre aisle. Qualification then has two jobs. Confirm that even the worst-case location remains inside the labelled range under the empty and loaded states written in the protocol. Record the coordinates so that routine monitoring probes are not all clustered in the easiest air.

Read contour plots and time-temperature traces together. A contour without the trace hides cycling. A trace without the floor plan hides whether the probe sat in a dead zone or in the airflow. Acceptance written as “all sensors between 2 °C and 8 °C” is incomplete unless it also states logging interval, door-opening challenge, how many sensors may be excluded, and the documented reason. Worst-case is a location and a time window. MAP-210 spends classroom time naming both before anyone signs the report.

April 2026

7 min · Sglearninghubplus training desk

Why kinetic mean temperature is not a pass mark

Kinetic mean temperature, often spoken of as MKT, compresses a time-temperature series into one number using the Arrhenius relationship. It is a useful way to compare the thermal load of a 14-hour delay at 12 °C with a 40-minute spike at 18 °C. It is not a substitute for the labelled storage range, and it is not a silent pass against a 2–8 °C specification. A shipment can show an MKT of 7.1 °C while spending two hours at 14 °C. Those two hours still need a product-specific decision.

The calculation also inherits every fault in the file. A logger started in an air-conditioned office, a clock set to the wrong zone, a probe taped to the outer carton rather than the payload, or a 41-minute gap all change the number. Before anyone keys the formula, the investigation has to establish that the channel represents the product. EXC-130 treats MKT as a supporting annex after the evidence pack is complete, never as the first line of a disposition.

Public stability statements sometimes quote MKT or mean kinetic temperature as a way to discuss excursions. That language still belongs to the manufacturer or the responsible person who holds the stability file. A warehouse or 3PL in Singapore can calculate MKT from a validated spreadsheet and send it with the logger graph. The calculation does not authorise release. Write the peak, the duration outside the band, the pack-out, and the product identifier. Then ask for a decision. The number without the graph is how guesswork gets a decimal place.

June 2026

9 min · Sglearninghubplus training desk

Qualifying a passive shipper for a 42-hour lane

A 42-hour lane from a Singapore warehouse through an airport and a receiving dock is long enough for a passive shipper to fail in ordinary weather. Qualification starts with a written payload, a labelled range, and an ambient profile that includes apron time, not only the hours in a 15 °C packing room. Catalogue “up to 72 hours” figures are usually generated against a laboratory cycle. They are a starting point for protocol design, not evidence that your SKU will arrive at 5 °C after a delayed ULD.

The protocol should state empty and representative payload, coolant conditioning, pack-out photographs, logger placement in the payload rather than in the lid void, and the pass criterion against the labelled band for the whole duration plus a justified margin. Summer and a cooler monsoon profile both belong in the plan if the lane runs year-round. Door or lid openings during security inspection need a challenge if that is how the lane actually works. PSL-120 uses 42 hours because it forces the conversation about hold time after a missed connection.

Operational qualification on the real lane still needs a logger file from goods-out to goods-in, with hand-off times that match the air waybill and the receiving stamp. If the qualified duration is 36 hours and the product sat 41 hours, the file is an excursion even if the graph stayed inside 2–8 °C by luck of a mild night. Requalification is due when payload, coolant brand, pack-out, airport handling or SKU changes, including an October product after a March pass.

August 2026

8 min · Sglearninghubplus training desk

Dry ice, UN1845 and the paperwork that stops a pallet

Solid carbon dioxide shipped as dry ice is UN1845. For air freight the quantity, marks, ventilation of the pack, and the statements on the air waybill are as capable of stopping a pallet as a failed temperature graph. Acceptance staff will look for the net mass of dry ice, the proper shipping name, and whether the pack can vent. A GDP warehouse that treats dry ice as “just coolant” finds out at the cargo terminal that the shipment is also dangerous goods.

IATA Temperature Control Regulations sit beside the dangerous-goods rules. Booking a 2–8 °C product in an active container is a different service from a passive pallet with a UN1845 charge for a frozen IMP. Mixing the two in the special-handling field is how a ULD is built for the wrong set-point. AIR-150 walks the booking language, the marks on the outer pack, and the conversation with the airline when the dry-ice mass changes after a re-ice at the origin warehouse.

Ventilation and hold time also belong in the GDP file. Dry ice sublimates. A pack that left Changi with enough charge for twelve hours will not cover a 26-hour tarmac delay unless the protocol said so and the receiving site knows how to inspect remaining ice. Write the mass at pack-out, the expected sublimation, and the inspection at destination. If the product is also a clinical material, chain-of-identity stays on the same page as the UN1845 mass. The usual stop is a missing net quantity or a pack that cannot vent.

October 2026

8 min · Sglearninghubplus training desk

Closing an excursion investigation without guessing

An excursion file closes when a named person records a disposition against evidence. It does not close when someone says the product “looks fine” or when kinetic mean temperature falls under 8 °C. Start with the logger: configuration, calibration due date, time zone, start delay, and whether the probe was in the payload. Add pack-out photographs, the air waybill or collection note, and every timestamp from warehouse door to receiving dock. If a 41-minute gap exists, write it as a data-integrity finding before you interpret the rest of the curve.

Then describe the thermal event in product language: labelled range, peak, duration outside the band, and whether the pack was opened. Ask the owner of the stability or technical file for a decision in a form that cites those facts. Local QA in a 3PL may quarantine and notify. Release of a medicine or an IMP is not a warehouse courtesy. CAPA should change something that failed: pack-out, lane duration, alarm delay, or the on-call roster. Restating “staff will be careful” leaves the next night shift with the same panel.

Teaching scenarios at Sglearninghubplus use a tarmac delay, a 03:40 compressor failure, and a gapped logger because each one fails a different part of the evidence pack. The tarmac case is usually a lane and booking problem. The compressor case is mapping, alarm and after-hours ownership. The gap is configuration and retrieval. Close the one that actually happened. EXC-130 marks a script that guesses the product is stable as an incomplete investigation, even when the guess later proves lucky.

Bring a file to class

If a mapping report or a logger gap is already on your desk, name it in the enquiry (without product identities you cannot share) and we will say which 2026 code fits.

Enquire Course catalogue