Tools / AMI teaching explorer

AMI teaching explorer

See how a 0–100 field score could combine four inputs, what each input adds, and when a score is shown without a status. The inputs are invented for teaching, and the weights are not AgriNova's formula.

The explorer recalculates with JavaScript, which is off in this browser. The page shows the example field's later week: its four inputs, the score and how it is built.

Examples

Example field

Teaching scenarios

From the earlier week's 75, the soil-moisture and water-stress conditions moved most, to 45 and 42.
From the earlier week's 75, the temperature condition moved most, to 54.
From the earlier week's 75, the soil-moisture condition, the water-stress condition and crop vigour moved most, to 30, 36 and 40.

These inputs are invented for teaching. The demo's North Field drops by the same 11 points, but the demo has no input values for it, and nothing here explains why North Field changed.

Four inputs, each from 0 to 100

Each input is a favourable-condition score: 100 is the most favourable condition and 0 the least. Temperature and water stress are turned around before they are combined, so a higher number always means less stress.

Inputs: Example field, later week (Sample: invented for teaching)

Crop vigour

Soil-moisture condition

Temperature condition

Water-stress condition

Teaching weights — not AgriNova's formula. Each input counts for a quarter (0.25) of the score, and the four weights add up to 1. You cannot change them here. How AgriNova AI would weight and calibrate the inputs for each crop and region is still being researched.

Evidence state

A status is shown only when the observation is usable, no more than 14 days old and its inputs agree. Otherwise the last valid score is shown and the status is withheld.

Fresh Usable, and no more than 14 days old.Stale The newest usable observation is more than 14 days old.Missing input One of the four inputs could not be measured.Conflicting The inputs disagree with each other.

Sample · Illustrative calculation — your inputs

Score

Score from the four inputs

64/100

Status: Review
Change from the earlier week: Down 11 points (75 to 64)

How the score is built

Sample · synthetic data

Each input times its teaching weight; the four contributions add up to the score
Input	Value	Weight	Contribution
Crop vigour	58	0.25	14.5
Soil-moisture condition	70	0.25	17.5
Temperature condition	58	0.25	14.5
Water-stress condition	70	0.25	17.5
Score (the sum, as a whole number)			64

Contributions are shown to one decimal and the score as a whole number, each rounded half away from zero; the score is rounded once, from the unrounded sum.

Bands

illustrative bands, not AgriNova's thresholds

Stable 80–100
Watch 70–79
Review 0–69

What the score is not: not a yield forecast, not a diagnosis, not an instruction. A change of 11 points is not an 11% yield loss.

The three demo fields, ranked by change

Sample · synthetic data

The guided demo's own scores, separate from the teaching inputs above: nothing you enter here moves them, and the demo has no input values for these fields.

North Field · Wheat64 AMI · Down 11 pointsReview
West Field · Canola71 AMI · Down 4 pointsWatch
East Field · Barley82 AMI · No changeStable

See the evidence behind each field

About this tool

How the explorer works, its cautions and its limits

What it is

A teaching tool. It shows how a 0–100 score could combine four favourable-condition inputs with fixed teaching weights, and how one evidence rule decides whether a status can be shown. It is not AgriNova AI's scoring method, which is still being researched.

The formula

Score = 0.25 × crop vigour + 0.25 × soil-moisture condition + 0.25 × temperature condition + 0.25 × water-stress condition, shown as a whole number. Change = score − the earlier week's score (75).

Cautions about a score like this

Inputs are converted to favourable-condition scores, where higher is better, before they are combined; a raw stress reading would otherwise pull the score the wrong way.
Soil-moisture and water-stress signals can overlap, so adding both can count the same dryness twice.
Data quality is kept separate from field condition: a stale, missing or conflicting observation changes whether a status is shown, never the inputs themselves.
No accuracy percentage is given: the teaching weights have not been tested against field results.

The one evidence rule

A score is shown only if it rests on a complete, usable observation.
A status, and the change, are shown only when the observation is usable, no more than 14 days old and its inputs agree.
Otherwise the most recent valid score is shown, labelled "Last valid score", the status is withheld and the evidence needed is named.

What the labels mean

Sample: The examples and scenarios are invented for teaching; they are not inputs of the demo fields.
Your input: A value you typed or set with a slider. It never leaves your browser.
Calculated: The score, its contributions and the change, from the formula above.

Limits

The demo fields have no input values; the explorer never shows any for them.
The bands are teaching bands taken from the demo's scores (82 Stable, 71 Watch, 64 Review), not AgriNova's thresholds.
Nothing is saved or sent: reloading the page or pressing Reset restores the example field's later week. Print the page or save it as a PDF to keep it.

Last reviewed: 2026-10-08

Next step

A score is a reason to look at a field, not an answer; the scouting checklist prepares a field visit that compares the area that looks different with a typical one.

Prepare a field visit with the scouting checklist