Book 2 / Parts 5–7 — Sensory Science, Professional Operations & The Future

The first half of Book 2 explains what happens inside the bean. The second half answers a harder question: how do you prove it, scale it, and keep it consistent batch after batch?

Once you understand the chemistry and physics of a roast, the work shifts. It is no longer about what is happening to the coffee — it is about measuring it, tasting it objectively, repeating it reliably, and building a business on top of it. This is the territory of Book 2's final three parts: Sensory Science (Part 5), Professional Operations (Part 6), and Looking Forward (Part 7).

These ten chapters are the bridge between the science and the craft. They take the working roaster from the cupping table to the production floor, from a single good batch to a process that delivers that batch every time, and finally to the technologies and pressures reshaping the trade. Here is a tour of what they cover.

Part 5 · Chapters 23–24

Cupping: a universal language for taste

Cupping, the book argues, is like wine tasting with a rulebook. A taster in Addis Ababa, Melbourne, or Brooklyn follows the same ritual and produces scores that mean the same thing. That standardisation — not the slurping — is what gives the data its power. Samples are roasted to a light-medium target (Agtron 58–63 ground), rested 8–24 hours, and brewed at a fixed 8.25 g per 150 ml with 93°C water, steeped four minutes before the crust is broken.

The 100-point SCA sheet then breaks the cup into ten attributes — fragrance/aroma, flavour, aftertaste, acidity, body, balance, uniformity, clean cup, sweetness, and overall. Each tells part of the story; the total reveals the whole. For a roaster, this is not ceremony. It is the most direct feedback loop you have: a way to stop guessing at profile adjustments and start making decisions grounded in sensory evidence.

Part 5 · Chapters 25–26

Triangulation: proving a difference is real

Here is a problem every roaster faces: you roast the same coffee at 18% and 20% development time ratio and you are certain they taste different. But are they, really? Triangulation answers objectively. You present three cups — two identical, one different — and ask the taster to find the odd one out. Guessing blind gives a 33% success rate. Do it consistently across eight sets and you have proof.

The statistics are unforgiving in the best way: getting six of eight correct means the odds you simply got lucky are under 2% — comfortably past the scientific significance threshold. This removes ego from the equation. If a panel cannot reliably tell two of your profiles apart, the difference is not perceptually real, however convinced you were. Threshold testing extends the same rigour to sensitivity, calibrating a palate like tuning a radio to catch a weaker signal.

Chapter 26 then connects those perceptions to chemistry. Coffee flavour, the book suggests, is an orchestra of hundreds of compounds — 2-furfurylthiol giving that roasted, coffee-smells-like-coffee note at sub-ppb levels; diacetyl lending butterscotch that fades over days of degassing; guaiacol marking the fine line between good smoke and ashy. Understanding which compound plays which note is what turns button-pushing into conducting.

Part 6 · Chapters 27–31

From one good batch to a reliable process

Production roasting, Book 2 says, is the jump from cooking dinner for friends to running a restaurant kitchen. A 5 kg batch responds to heat in seconds; a 25 kg batch lags, and the physics changes. The discipline that makes this work is measurement. Three numbers on every batch are non-negotiable: Agtron colour (your immediate feedback), weight-loss percentage (your early warning that the green has shifted), and development time (your flavour-progression check). After ten to fifteen solid batches you calculate the mean and standard deviation, then set warning limits at ±2 SD and control limits at ±3 SD — borrowing statistical process control straight from manufacturing.

Blending gets the same blend of theory and tasting. The book's three-layer architecture — a 55–70% base for body and chocolate, a 20–30% accent for fruit and florals, a 5–15% seasoning that customers cannot quite name — is designed on paper but proven on the cupping table, because flavours never add linearly. Sample roasting, covered in Chapter 29, is the audition process: a standardised small roast that lets you compare five Ethiopian lots fairly before committing real money.

And none of it survives without records and safety. Roasting by feel, the book notes, is like navigating by the stars — it works, but you cannot retrace your route or explain it to anyone else. Logging software (Artisan, Cropster, RoastLog, Typica) is the GPS that makes invisible data actionable. Chapter 31 is blunt about the stakes on the floor: chaff is the primary fire risk, and a professional roastery is a factory producing heat, smoke, and mechanical motion. Daily chaff emptying, weekly duct cleaning, and a rehearsed fire response are foundational, not optional.

Part 7 · Chapter 32

Where roasting is heading

A generation ago, roasters worked with thermometers, stopwatches, and intuition. Today there is real-time curve analysis, cloud-connected sensors, and AI suggesting profiles from a green coffee's density and moisture scan. The book's closing chapter maps what is coming: computer-vision colour measurement reading beans live inside the drum, IoT monitoring for multi-site operations, and electric roasters that the author expects to become viable for small and mid-size roasteries by 2030.

Two pressures loom largest. Sustainability is moving from nicety to expectation — a mid-size roastery can burn 20,000–30,000 kWh a month, and carbon accounting is becoming standard. And climate change is already redrawing the map, pushing cultivation to higher altitudes, prompting new climate-resilient varieties, and shifting the cup character of familiar origins. Against all of it, the through-line is human: lighter roasts, single-origin transparency, and QR-code traceability all reward the roaster who marries the science to the craft rather than choosing between them.

Who this post is for

If you have moved past your first successful roasts and want to taste objectively, scale without losing your voice, and run a roastery that is consistent and safe, this is the half of Book 2 written for you. It is equally useful for the curious enthusiast who wants to understand what separates a professional operation from a home setup — and where the whole trade is heading next.

Written by Wing Yuen, SCA Authorised Trainer at Tasse Coffee Roastery in Shinjuku, Tokyo. The Roasting Bible is a two-book series drawn from years of roasting, teaching, and chasing the science behind a great cup.