Owners and managers

Setting up a trial

Five steps, one screen at a time. A bar across the top fills as you go and every step is tappable, so you can jump back to fix one field without walking the whole form again. Until you press Start, Setup is the only tab there is — there is nothing to log yet.

1 · The trial

A name, what you are trying to find out, and the question that shapes everything below it: what are you trialing?

1Genetics — you are comparing cultivars, seed lines or clones against each other. The cultivars are the comparison, so there is no treatment to invent: one table, one row per cultivar.
2Environment — you are comparing the conditions the crop is grown in, above ground or at the root. Light level, temperature split, CO2, humidity, irrigation, media, nutrition.
3Cultural practice — you are comparing something you do to the plants. Canopy management (pinching, topping, defoliation), plant density, cycle time.

Change it later if you got it wrong. Collapsing a trial down to Genetics removes the extra treatments, so it asks first.

2 · What you are comparing

On a genetics trial this is six plain questions, not a table:

Seed or clone? Asked once for the whole trial, not once per line.
Where did they come from? One breeder or supplier, applied to all of them.
How many cultivars? A plus and minus. Each new one inherits the answers above, so adding a sixth line does not mean retyping the breeder a sixth time.
What are their names? One box each.
How many plants of each? One number that sets them all; the running total is shown underneath.
Where are they going? Buttons for the rooms and zones mapped on the Facility page.

There is no Role field. In a genetics screen the cultivars are the comparison, so marking one an entry and another a control changed nothing anyone looked at.

On an environment or cultural trial there are two tables, because the genetics and the thing being tested are different questions.

Each treatment row is one column of the trial.

1What you are doing differently. One row per treatment. Your normal practice is a row too.
2Mark your normal practice as Control. Everything else is measured against it, and without one the results table has no baseline.
3How many plants are in that treatment at the start. With more than one cultivar a grid appears below so you can vary it per cultivar.
4Where they live — picked from the rooms and bays mapped on the Facility page, down to the valve.

Every cultivar crossed with every treatment makes a plot — the unit that carries plant counts, results and photos. Two cultivars against three feed rates is six plots. A five-line genetics screen is five plots, because its one treatment is held constant and kept out of your way.

3 · How long

Switch off any stage you are not running and set roughly how many days each takes. These are a plan, not a promise — the tracker records what actually happened and shows you the difference.

Set points are folded away under a button, because plenty of trials never need them. Open it and there is one shut row per variable — photoperiod, light, day and night temperature, humidity, CO2 — each already showing its answer, so you only open the one you want to change.

Every variable is asked the same question, and the answer is Same throughout unless you say otherwise:

1Same throughout. One number for the whole run. This is the default and it is the right answer most of the time.
2Changes. The number moves through the cycle. Tap it and the phases are laid out for you, one per stage, already carrying the number you typed — so you only fill in the ones that differ. Rename a phase, change its length, add one or remove one: splitting flower into weeks 1–3 and 4–8 is two taps. The day range beside each phase is worked out, never typed.
3By treatment. A different number for each treatment. This is how you say the variable is the trial — 22° against 26° — rather than a condition the trial happens to run under. It is only offered when there is more than one treatment, it marks the row as the trial variable, and it is named on the report as the thing that differed.

The bar above the log shows whichever number applies today, so a plan that changes mid-flower shows what the crop is meant to be held at now rather than an average of the stage.

4 · What you will measure

Starts empty on purpose. Add only what you want recorded every single visit. Every box here is a box somebody fills in on a phone in a hot greenhouse, and a long form is a form people quietly stop filling in. Two or three is usually plenty.

The section is laid out as a walk through the stages you switched on in section 3, in the order the plants go through them. At each one there are two lists, because “measured in propagation” was one answer to two different questions.

Most of what a trial is run for is taken as plants LEAVE a stage, which is why the app says out loud what the end of each one means:

1End of propagation — the cuttings come out of the dome. Rooting success, root initials.
2End of veg — the plants go into flower.
3End of flowering — the day they come down. That date is the harvest date: the move dialog says so, and it is what every yield is dated from. Fresh weight is taken here.
4End of dry — out of the dry room. Dry weight.
5End of post-harvest — bucked weight, then flower, smalls and trim, then whatever the lab says. Post-harvest is on by default, because that is where most of the numbers live. Every stage starts empty: you add the measurements you actually take.
6The last stage you switch on is the end of the trial, whichever one that is. Turn post-harvest off and the dry room becomes the finish; the block says so, and so does the button on the Log tab.

1Measured during the stage. Every check-in, on the schedule at the bottom of the section. Height and vigor live here.
2Measured at the end of the stage. Once, as the plants come out. Rooting success is the clean example — count it in week one of propagation and the answer is zero, which is not a result, it is a form filled in at the wrong time. Wet weight belongs to the end of flower, dry weight to the end of dry, and bucked, trimmed and the three grades each to the end of their own step.
3Both is a real answer. Height every week and height at the end is how you get a growth curve with a final number on it. Tap it in both rows.
4Nothing is a real answer too. A stage with nothing ticked reads N/A and is finished, not unfinished. Photos and notes still apply there.
5Nothing follows the stages. The last stage you switch on is the end of the trial. Yield, potency and the rest are measured at the end of post-harvest like anything else.

Neither row locks anything. The end-of-stage measurements are folded away in the log entry rather than hidden — one tap, on any day, and a rooting count can be typed in on a Tuesday if that is when it got done. What the split changes is what a grower is asked for at a normal check-in, and what the app says on the way out of a stage: move plants on with an end-of-stage measurement still empty and it tells you, before you confirm, while the plants are still there to measure.

Everything about a measurement is edited inside the stage it belongs to — its name, its unit, whether it is a number or a 1-5 rating, and how often it gets taken. Add one to propagation and its boxes appear in the propagation block, under your finger. They used to live in a single table folded away at the bottom of the section, below every stage and below harvest, which is why adding a measurement looked like it had done nothing at all.

1How often lives on the measurement. "every 3 days" sits in the row, because a rooting check every two days and a height every fortnight are different questions and were sharing one number. Leave the box empty and it uses the trial-wide fallback at the bottom of the section, which is what every existing trial does.
2The same measurement can be used at more than one stage. Height in veg and height in flower is one measurement, so it stays one column in the results. It shows a row in each stage and editing either one edits the measurement; the note under the rows says so when it is actually happening.
3The × means "not here", not "delete". Removing height from flower leaves it in veg. Take it off the last stage it is used at and the app asks before dropping it from the trial — and whatever was already logged against it stays in those entries either way.
4Already defined somewhere else? A short row of chips at the bottom of each stage offers the measurements this trial already has, so you pick one up rather than typing it again and ending up with two columns that mean the same thing.

A grower is then only asked for the measurements that belong to the stage the plants are in. Change the stage on a log entry and the measurements change with it.

Overdue is a property of a measurement now, not of the trial. Leave a measurement's every N days empty and it is never chased; fill it in and it is. On the Log tab each one carries its own schedule on its chip — "Vigor 1-5, every 2d, 7d late" — and the dashboard counts a trial as needing attention when any measurement at the stage the plants are in has run past its own interval.

There is no photo reminder any more. It was one interval for the whole trial, set in the same block, and it went with it.

5 · Review and start

Everything optional lives here — trial code, replicates, density, pot size, media, notes — along with the button the whole form leads to. Press Start when the plants are actually in. That timestamp is what every week number and stage is counted from, which is why the Log tab does not appear until you do.

Changing a trial after it has started

Every step above stays editable. Add a measurement in week six and it appears on the next log entry — nothing is frozen by pressing Start.

Save changes sits at the bottom of each step once a trial is running. Setup writes as you go and always has, so nothing is riding on that button — what it does is stop waiting, push the change up now, and say Saved. Without it the only evidence your edit had landed was reloading the app to see whether it came back, which is a horrible way to find out.