Skip to content

Overview

Making a field plan a farmer can read in one look

Verge plans the route a tractor drives across a field. The maths was solid. What a farmer met on screen was every control at once.

1 / 3 · Setup

Verge path planner prototype: Setup
Verge path planner prototype: Optimise
Verge path planner prototype: Results
Swipe through the screens.

About the project

Straight and curved planning showed different controls, everything packed into one panel, and no route from the settings that matter to the many that rarely do.

There is also no best plan. Time, fuel, coverage, tracks and soil loss pull against each other, and every field trades them differently. Showing every control hands that maths to whoever is about to drive the tractor.

Services
Product Design
Industry
Agriculture and farm equipment
Engagement
Ongoing
Period
Jan 2026 – now

What we solved

A plan is generated by the time you arrive, balanced by default, with a few goal-based alternatives. The technical controls stay, as overrides.

Plan From got search, a preview of the field and the equipment, and a way to detach. Two saved plans can look identical and differ in ways that change the whole job.

The redesigned setup screen, with the operation picked from a row, equipment and headlands grouped, and the field carried on the map.
The previous setup screen: boundary, operation, swath width, track type and headland passes stacked as one long form.
Setup, before and after. Drag the handle to compare.
The redesigned results screen: the plan on the map, its settings in an accordion, and time, tracks and distance under it.
The previous curved track screen, with the steps for adding a terrace written out as a paragraph to follow in order.
Results, before and after. Drag the handle to compare.

How the results page got here

Five passes, from a control panel to something readable.

  1. A dense technical control surface

    Simple setup, then everything at once. People reached a plan and could not tell what it was worth or how to change it.

  2. Lead with a default, not analysis

    The early concepts still treated results as a place to compare and tune. We moved to showing the balanced plan by default, alternatives framed as goals — fastest, least soil loss — rather than settings.

  3. Goals early, refinements late

    What defines the plan moved to the start. What you can only judge by looking at the map — heading, headlands, tram lines, boundaries — stayed with the result.

  4. Results became somewhere you work

    A plan on arrival, answering as you change it instead of sending you back to regenerate. Plan From shows its reference alongside, and disconnects once the new plan has moved on.

  5. Contextual and progressive

    Strong default, immediate result, adjustments where you can see what they do, everything else disclosed when it matters.

Where we stand now

Forage harvest is in progress. Mowing, raking and harvesting generate together from the mowing setup; only mowing is editable in V1.

The constraints are real: screens as small as 600 by 500, Angular Material underneath, and up to a minute to process a big field.

Splitting fields into zones is still open: who owns the geometry, what a guidance line belongs to, what undo means afterwards.

What we built

  • Split UI

    Split a whole field into parts, or just the inner area under one shared headland. You choose before you draw, and it stays clear whether a change hits one part or all of them.

  • Drawing on maps and reference lines

    Draw, import, save and reuse an alignment line, or split a field with one. Drawing opens because you are doing something, not as a mode on a toolbar.

  • Results

    One balanced plan, settings in an accordion rather than a wall, and time, tracks and distance under the map.

  • Plan From

    Build on an existing plan, with search and a preview to tell similar ones apart. Carries the seasonal work too: mowing, raking and harvesting from one setup, with an operation selector on the map.