1.1 Evidence first

A drone sees the whole field at once. That is its gift. It is also a trap, because a picture this detailed tempts everyone to jump to a conclusion.

So KADS Field Utilities reports what the imagery shows, and stops there. It gives the grower numbers, maps and dates. It does not grade the field. It does not tell him what to do. The grower knows his ground, his costs and his risk. The decision is his.

When the numbers are thin, we say so. A single average for a whole field has nothing to compare against, so we never call it good or bad. It is a baseline inside that field, not a verdict on it.

1.2 Make a change, measure a change

Every practice change is a question. Did the extra nitrogen pay? Did the new variety stand better on the wet end? Did the fungicide pass change anything we can see?

Here is the cycle Field Utilities is built around:

1. Gather the data.
2. Analyze it the same way every time.
3. Report it plainly.
4. Invite the grower and the agronomist to plan the next operation together.
5. Measure how the crop reacts.

The last step is the one that matters most. A change you never measure is a hope. A change you measure becomes something you know.

The goal behind all of it is even yield across the entire field.

1.3 Keep the record

A field report is a snapshot. A season of reports is a story. Several seasons are evidence.

Field Utilities keeps the field's outline, its crops, its varieties, its planting dates, and every flight over it. Nothing is deleted. When an outline changes, the old one is kept with its date, so an old report still reads correctly years later. When the farm changes hands or tenants, the history stays with the ground.

1.4 Work together

The grower, the agronomist and the drone company all look at the same field record. The grower knows the ground. The agronomist knows the crop. The drone company knows the aircraft and the imagery. Each one sees something the others miss.

Field Utilities invites the agronomist to take part at the key operations of the season, so the plan and the measurement happen in the same place.

1.5 What KADS Field Utilities does not do

• It does not make decisions for the grower.
• It does not recommend a product, a rate or a practice.
• It does not replace walking the field. Every report carries this note: "This technology can introduce small variations with every image. Health indexes are approximations of ground truth. Manually scout troubling areas."

2.1 People and roles

Drone company. The business that flies the fields and owns the account. Each drone company's growers, fields and reports are private to that company. KADS is one drone company. The subscription belongs to the drone company.

Primary contact. The one person at a drone company, a grower or a farm owner who is billed. Each has exactly one.

Grower. The person or business that farms the field this season. A grower can log in and see his own fields.

Farm owner. The person or business that owns the ground. A farm owner may be a different party from the grower. When one business both owns and farms the ground, it is listed twice, once as each, so a later change of tenant does not rewrite history. Farm owners do not log in today.

Agronomist. The grower's crop advisor. The grower decides which fields an agronomist can see. Status: planned. See section 5.4.

Member. A person on file with a drone company, a grower or a farm owner. A member may have no login at all.

Role and title. A role says what a person can do. Today the roles are owner and member. A title, such as Pilot, Agronomist or Farm Manager, is only a label. A title never changes what a person can do.

Pilot. A title, not a role. A pilot flies and uses the tablet.

2.2 The ground

Parcel. A piece of land as it is recorded, usually by its county parcel number. A parcel can hold several fields. A field may have no parcel.

Field. A piece of ground that is farmed as one unit. The field is the ground itself, and it is permanent. A field name only has to be unique inside its parcel.

Outline. The line around a field, also called its boundary. An outline is dated. Redrawing an outline adds a new dated outline and keeps the old one.

Inside area. An outline drawn inside a field for a reason. Each inside area has a purpose:

Variety zone. Where one variety stops and the next starts inside a field. A variety zone gets its own crop, variety and planting date. A test plot is never a variety zone.

Season. One crop year on one field. The grower, the farm owner, the crop, the variety and the planting date belong to the season. The season year is the crop year, not the flight date. Wheat drilled in September 2025 is the 2026 crop.

2.3 Flights

Visit. One trip to one field on one day. The visit is the record. A correction to a visit is saved as a new version, and the old one is kept.

Flight purpose. Every visit records why it was flown:

• Scout: to read the crop.
• Map: to confirm or add an outline, add a feature, or gather elevations. A mapping flight skips the crop analysis.

Mission. A flight job over a grower's ground, planned and then flown.

Suggested flight. A flight Field Utilities thinks would show something useful, based on the crop, its stage and the weather. Only the drone company sees suggested flights.

Verification mission. A set of calibration flights, A through F, flown to measure how one aircraft actually performs. It is not a mission over a grower's field.

Tender. A spot where the drone trailer parks to refill and swap batteries.

2.4 Imagery

RGB. An ordinary color photograph of the field. It is a picture, not a measurement.

Multispectral. Images taken in several narrow bands of light, including bands the eye cannot see. The health indexes are built from these.

Orthomosaic. Hundreds of overlapping images stitched into one map of the field, corrected so distances are true.

Calibration panel. A card of known reflectance photographed before and after the flight, so the camera's readings can be corrected for the light that day.

Pixel size (GSD). How much ground one pixel covers. A 4 inch GSD means one pixel covers 4 inches of your field.

2.5 Health indexes

NDVI reports how much green plant is in the field. It is the better index early in the season, while there is still bare ground between the rows. Once the canopy closes, NDVI stops separating good from better.

NDRE reports how hard the plants are working. It keeps responding after the canopy closes.

Every report prints both all season, because it is useful to scout the areas where NDVI and NDRE report different colors.

2.6 Analysis

Mean. The average index value over an area.

Standard deviation. How spread out the values are around the mean. A small number means the area is even. A large number means it is patchy.

Zone. The field cut into five groups of similar values at natural breaks. Zone 1 is the weakest and prints red. The breaks are set for that field on that day only, so a zone number never transfers to another field or another flight. The smallest zone mapped is a quarter acre.

Canopy. Crop height times cover. It is a stand-in for how much plant is there. It is not a weight.

Observation. A spot worth noting: a weed patch, ponding, a tile riser, hail damage. See section 2.8.

2.7 Experiments

Treatment. The thing done differently: a rate, a product, a variety, a timing.

Test plot. The ground that got the treatment.

Check strip. The ground that did not. Also called the untreated area or the control.

Replication. Doing the same comparison in more than one place in the field, so one wet corner cannot decide the answer.

2.8 Observations

An observation starts as a candidate. A person then confirms it or dismisses it. A dismissed spot is not raised again.

• Perishable observations belong to the day they were seen: a weed patch, lodging, hail damage.
• Durable observations stay with the ground: a tile riser, a waterway, a culvert. They are carried to the next flight and confirmed rather than entered again.
• A durable feature that is gone, such as a filled waterway, is retired with a date. It stays in the record.

2.9 Files

KML. A small map file that holds an outline. Most farm software can export one.

Field report. The PDF built from one visit.

Intake record. What the pilot entered on the tablet at the field.

4.1 Start with one question

Write it down in one sentence. "Does 30 more pounds of nitrogen raise NDRE on the east half?" is a question. "Is the field healthy?" is not.

One question per experiment. Two changes in the same strip give you one answer and no way to know which change caused it.

4.2 Design the test

• Put a check strip beside every test plot. The check strip gets everything the test plot gets except the one change.
• Cross the field. Ideally, the test plot is split across the field to take in both high and low ground.
• Repeat it. Two or three pairs across the field beat one pair. Up to four test plots and a check strip can be compared on one report.
• Keep the ends out. Headlands and turn rows get different treatment from the machine. Keep plots off them where you can.

4.3 Record the test in Field Utilities

1. On My Fields, open the field.
2. Choose Mark an area inside this field.
3. Give the area a name, pick its purpose, add a note, and draw it.
4. Send it. Your drone company reviews it before it is used.

Mark the check strip the same way, with the purpose Check strip, left untreated. Use Nitrogen reference strip for a strip given plenty of nitrogen.

Do not mark a test plot as a variety split. A variety split changes how the whole field is divided.

4.4 Plan the flights

• Fly before the treatment and after it. A single flight after the fact shows a difference but not a change.
• Fly at the growth stages that matter for the question. Your drone company's suggested flights list does this timing for them.
• Keep flights comparable: the same height, the same time of day where possible, the same processing.

4.5 Read the results

• Compare the test plot to its own check strip, on the same flight.
• Compare flight to flight in the same season. Comparisons across seasons are not made.
• Look at the standard deviation as well as the mean. A treatment can raise the average and make the plot patchier.
• Two check strips that disagree are never averaged. The disagreement is the finding.

4.6 Close the loop with yield

The imagery measures the crop during the season. The combine measures it at the end. A planned module will compare yield monitor data against the field reports.

4.7 Common mistakes

• One test plot and no repeat.
• No check strip.
• Two changes in the same plot.
• Flights at very different times of day.
• A check strip that got the treatment by accident on the headland.