From 1,140 acres to 23 places worth a walk.

The operation covers about 1,140 acres: winter wheat in five blocks strung along eight miles of road, and avocado on terraced slopes. One field agronomist from the regional input distributor advises this grower alongside many others. An independent pest scout walks the orchard.

• Five wheat blocks, miles apart, mostly judged from the field edge between other jobs.
• An orchard with more rows than a scout can walk in a week.
• Problems were found when someone happened to pass them, not when they started.

Infrag flew five scans between December 2025 and April 2026, each timed to a decision. The first, over 500 acres of wheat after emergence, was free. The grower ordered the other four. Each time, the agronomist and the scout opened the map before going out, asked the AI Advisor about what looked off, marked the suspected areas and walked to those first.

• Three wheat scans: after emergence, before the herbicide decision, and after the pass.
• Two orchard scans, in February and April, compared side by side.

Each finding below was a suspicion on the map first. It became a fact only when someone checked it on the ground.

1. Winter wheat · December scan, after emergence

   Draft — sample data

   What the map showed

   Thin, uneven strips along the eastern edge of two blocks. The agronomist marked four of them as suspected emergence gaps.

   What the ground check found

   Three were gaps where the seed drill had run light. The fourth was a wet corner that had filled in by January.

   What the adviser did

   Nothing was reseeded: the agronomist judged it too late in the window. The drill rows were noted for the grower's pre-season service.

2. Winter wheat · January scan, before the herbicide decision

   Draft — sample data

   What the map showed

   A patch in the northern block with a different color and texture from the wheat around it. Asked about it, the AI Advisor outlined it as a suspected broadleaf weed patch and, on request, marked thirteen similar areas across the five blocks.

   What the ground check found

   The agronomist walked to all fourteen. Eleven were broadleaf weeds. Three were not: two were bare, stony ground and one was animal damage.

   What the adviser did

   The eleven confirmed patches sat in two of the five blocks. The agronomist recommended a herbicide pass on those two blocks and none on the other three. The March scan, set beside January's, showed the grower what had changed.

3. Avocado · February and April scans, compared

   Draft — sample data

   What the map showed

   With the two scans side by side, five places stood out where trees had thinner canopy than their neighbors. The pest scout marked each one as a suspected problem.

   What the ground check found

   Three were real: two rows fed by a blocked irrigation line, and a group of trees standing in wet ground at the foot of a slope. Two were younger replants, smaller only because of their age.

   What the adviser did

   The line was cleared that week. The scout asked for the low corner to be scanned again after drainage work, to compare.

There is no yield figure here. These numbers describe what was looked at, how often a suspicion held up, and whether the grower came back.