PRF3Non-mechanical robotic weeding

What it actually is

One thing before anything else, because it changes who this page is for. This action sits in the SFI 2024 expanded offer, and the SFI 2026 offer that followed it does not include it: GOV.UK have taken the action's own page down, so its wording and rate now stand in the SFI 2024 actions print version, its compatibility lists name no SFI26 actions, and it does not appear in GOV.UK's funding finder listing of the current SFI offer, which carries only its mechanical sibling PRF4. So if you hold PRF3 in an existing agreement, this page is for you for the rest of that agreement's term; if you are planning a new application, this is not an action you can add to it, and the robotic weeding action on offer is PRF4. Check your own agreement, and treat your agreement documents as the authority on what you hold.

Source: GOV.UK, SFI 2024 actions print version, page 312, and GOV.UK's funding finder listing of the current SFI offer, both checked 23/08/2026.

The action itself: weeds in the entered crop are controlled by robotic equipment that identifies and locates them, then destroys them with a laser or high voltage electricity. A pre-emergence residual herbicide is allowed on the chosen crop at or within days of its establishment; beyond that, you must not use any other herbicides on land entered into this action, which is wider than the crop and wider than the growing season. GOV.UK's stated aim is weed control with minimal herbicide use, supporting productivity, water and soil quality, biodiversity and integrated pest management.

Source: GOV.UK, SFI 2024 actions print version, page 312, checked 23/08/2026.

What it pays, and what that comes to

The rate is in the panel at the top of this page. Multiplied out:

Source: arithmetic on the payment rate shown at the top of this page, which comes from Tilth's copy of the GOV.UK action catalogue and carries its own last-checked date. The 20 hectares is an illustration.

Is it worth it on your ground?

For an existing holder the question is settled; for anyone else it cannot arise, so this section is short. Like the other precision actions, the payment was never sized to buy the technology: laser and electric weeding units are six-figure machines or contractor services with rates to match, and the per-hectare payment is a contribution towards using one, not a route into owning one. What the herbicide-free requirement saves in chemical and application costs, and what the robot costs to run across your hectares, are both your figures, and the sum lives or dies on them.

Source: the payment rate at the top of this page. The judgement is Tilth's own reading rather than scheme guidance.

Does it work, and what makes it work better

This section is what we have found in published sources about the practice this action pays for. It is not scheme guidance, it is not Defra's or the RPA's view, and it adds no requirement beyond what the GOV.UK action page sets out. We have named every source and given the date we checked it. What it means for your ground is your decision.

Robotic weeding is too new for the graded evidence assessments this site leans on elsewhere, so what we can offer is a recent peer-reviewed trial, and it happens to test exactly this action's technology. In 2024, three research trials in New Jersey and New York state compared laser weeding using a commercial unit against pre and post emergence herbicides at label rate in beet, spinach and pea. Laser weeding was as effective as or superior to the S-metolachlor, bentazon and phenmedipham herbicides in controlling erect annual weeds, including common lambsquarters and common ragweed. Against untreated controls it reduced weed cover by at least 45 per cent and density by at least 66 per cent, with at least 97 per cent less weed biomass by season's end. Crop stunting did not exceed 1 per cent, and crop biomass increased by at least 30 per cent where laser weeding replaced the herbicide programmes. The same trials found the weakness: laser weeding was less effective on purslane, because it keeps emerging in sequence, and on annual grasses, whose growing points sit protected below ground.

Source: Sosnoskie and others (2025), Deep learning-based laser weed control compared to conventional herbicide application across three vegetable production systems, Pest Management Science, checked 23/08/2026.

The caveats travel with it: one season, one country, three vegetable crops, one commercial unit, and the authors' own conclusion that further work is needed across environments and weed species. It is a promising trial, not a settled evidence base, and nothing in it measured PRF3 itself.

Source: the conclusions of the same paper, checked 23/08/2026. The framing is Tilth's own reading.

What you actually have to do

Use robotic equipment that identifies and locates weeds and destroys them with a laser or high voltage electricity, on the entered crop. A pre-emergence residual herbicide is permitted on the chosen crop at or within days of establishment; beyond that, the print version's wording is that you must not use any other herbicides on land entered into this action, which is wider than the crop and wider than the growing season. On arable land and non-permanent horticultural crops the action is rotational and can move with the cropping; on permanent horticultural crops it is static.

Source: GOV.UK, SFI 2024 actions print version, page 312, checked 23/08/2026.

Things worth watching

This section is Tilth's own reading of where this action is easy to get wrong. It is not scheme guidance, it is not sourced from Defra or the RPA, and none of it adds a requirement beyond what the GOV.UK action page sets out.

The herbicide window has one door. A pre-emergence residual on the chosen crop at establishment, then no other herbicide anywhere on the entered land, not just in the standing crop. A weed problem the robot cannot solve mid-season has no chemical fallback inside the action, and the season plan needs to reflect that before the crop is in the ground.

The technology is the agreement's hostage. Three years of the action means three years of access to a laser or electric weeding machine, owned or hired. A contractor exiting the trade, or a machine that proves wrong for your crops, does not change what the agreement requires.

Rotational, except where it is not. On arable and non-permanent horticultural crops the action moves with the rotation; on permanent horticultural crops it is static. A holding with both kinds carries both rules at once.

Keeping the evidence

Also Tilth's own reading. GOV.UK sets the requirements; what follows is a practical suggestion about record keeping.

GOV.UK asks for the job data from the weeder or tractor recording system, identifying the machine, the dates and the areas, field operations records with parcel references and sowing and harvest dates, and contractor invoices where the work is hired in. The machine's own log is the backbone here, which makes this one of the few actions where the evidence mostly writes itself; the job is keeping it, filed against the right parcels.

Source: evidence requirements, GOV.UK, SFI 2024 actions print version, page 312, checked 23/08/2026. The suggestion is Tilth's own.

The part only your own ground can answer

For the farms that hold this action, the running question is what the robot actually saved: the herbicide bill it removed, the passes it replaced, the weeds it missed. That comparison lives in your spray records and operations log, side by side, which is where Tilth keeps them. Request an invite.