Fordo journal
5 min read · September 2, 2026
Every batch we roast — on the Giesen W6A in Magelang and the Probat P25 in Sentolo — gets logged in Artisan, the roast-curve software running on the monitor next to the drum. The ticker crawling across our homepage is a row from that log, and it isn’t decoration: Batch 4092, Gayo natural, charge 214.3°C, dev 24.1%, drop 208.0°C, done in 11:42.
Those numbers decide what ends up in your cup, so it’s worth knowing how to read them. Here’s the whole row, field by field.
The batch number ties a physical bag back to one roast session. Tell us a batch number and we can pull up its whole curve, its cupping result and the green lot it came from — that’s the traceability our wholesale customers rely on, and it starts with this field.
Origin and process are the shorthand for what went into the drum: Batch 4092 was a Gayo natural, the slow-dry 300H lot from Bener Meriah.
Charge is the temperature the drum is holding when the beans go in — 214.3°C for this batch. It sets the energy of the whole first phase: charge too hot and the bean surface scorches before the inside catches up; too cool and the roast drags, flattening the aromatics before they ever form.
Charge is also why batch size matters. Six kilos on the W6A and twenty-five on the P25 soak up heat very differently, so “the same profile” on two machines means matching what the bean experiences — not copying one number across.
Between charge and drop, Artisan plots ROR — rate of rise, how many degrees the bean temperature climbs per minute. A healthy roast has an ROR that falls smoothly through the whole session; when it stalls flat or spikes, the cup pays for it later as papery or harsh flavors, and the log shows exactly where it happened.
Somewhere past the halfway mark comes first crack: the beans audibly pop as steam breaks their structure, like slow popcorn. It’s the single most important timestamp in the log, because everything after it is called development — and development is where the roast’s sweetness and balance are decided.
Dev% is the share of total roast time that happens after first crack. Batch 4092 ran 24.1% — of its 11:42 total, the last 2:49 or so was development.
Cut development too short and the coffee tastes grassy and sharp, sweetness never fully formed. Stretch it too long and you trade aromatics for flatness, sliding toward roasty and dull. Every lot has its window, and dev% is how we prove each batch stayed inside it.
Drop is the bean temperature at the moment the door opens and the batch spills into the cooling tray — 208.0°C here, a filter-appropriate exit. Together with duration (11:42), it fixes where the roast landed on the light-to-dark line: same drop reached much faster or slower is a different coffee.
Moisture — 11.4% — is the odd one out, because it’s measured before the roast ever starts. Green beans arrive at our intake and get a moisture check as part of QA; beans outside the safe range dry out or roast unevenly, and the fix is rejecting the lot, not adjusting the curve.
The last field in the row is the only one a machine can’t fill in. A curve can sit perfectly inside its reference and the cup can still be wrong — so every batch goes to the cupping table within 24 hours of roasting, and only a batch that passes there gets packed. The log tells us what the machine did; the QC field records whether a human agreed.
If you want to taste what a passing grade tastes like, Batch 4092’s lot is on the shelf.