Open Source RoboticsRobots, software & learning

OpenWeedLocator

Robot / hardware build

Raspberry Pi camera and relay subsystem for targeted weed treatment, with printable enclosures and configuration-specific wiring requirements. [1]

Build & first use →Official website

For
Agricultural robotics developers and builders evaluating camera-triggered tools [1]
Cost
No current cost or bill of materials established yet.
Source rights
Named licenses are documented by part. Review licenses
Unknowns
tools not fully established. Review gaps

Source researched · Not tested by us
Sources checked · Research details

Upstream photograph of an OpenWeedLocator installation on a larger agricultural vehicle; the vehicle is separate equipment.
Upstream photograph of an OpenWeedLocator installation on a larger agricultural vehicle; the vehicle is separate equipment. Guy Coleman and OpenWeedLocator contributors · MIT. Original image unchanged; displayed responsively. Preview re-encoded for faster display; the link opens the original file.
Licenses & related hardware

Requirements · Research details and dates

On this page

Overview

OpenWeedLocator, or OWL, uses a Raspberry Pi camera to detect vegetation and drive relay-controlled outputs. This profile concentrates on the Compact OWL enclosure and the color-based detection route. The vehicle, spray plumbing or other treatment mechanism is a separate integration; OWL source availability does not make a complete agricultural robot. The current source and documentation were preserved as separate revisions. [5] [7]

Is this for you?

Known limitations
  • Color-based detection does not distinguish green crop from green weeds
  • Connector naming differs between Compact BOM and wiring instructions
  • No complete vehicle, treatment system or delivered budget established

Scope and source-reconciliation findings.

[5] [7] [10] · Editorial assessment

What you need

Operating system
  • Raspberry Pi OS 64-bit

[8] [9]

Host computer

Selected Compact OWL: Raspberry Pi 5 4 GB; docs also list Pi 4 and 3B+

[7]

Dependencies
  • Camera cable matched to the Pi, SD card and writable storage for recording/controller modes
  • Printed enclosure, mounts, inserts and seals
  • Driver board or separate relay HAT and 5 V regulator
  • External 12 V supply and selected treatment equipment

[7] [8] [12]

Hardware & parts

Hardware configuration

Camera and actuation subsystem; vehicle and treatment equipment supplied separately

[5]

Sensors
  • Selected camera route: Raspberry Pi Global Shutter Camera with suitable CS-mount lens

[7]

Assembly & first use

Project documentation [1]

Cost

No current, complete configuration-specific build cost has been established for this profile.

3D preview

Preview not available yet. Mesh files were identified. Their source inventory does not by itself establish assembly hierarchy or component placement.

Project notes

Select one complete hardware arrangement

The Compact guide lists a Raspberry Pi 5 with 4 GB RAM, with Pi 4 and 3B+ alternatives, and several camera choices. Its recommended Global Shutter Camera needs a suitable lens, camera cable and mount. A printed enclosure also needs inserts, screws, sealing cord and a 37 mm lens filter. Printed plastic alone does not complete the enclosure, and no environmental rating was verified here. [7]

Power and outputs have two routes: the official OWL driver board combines functions, while a generic relay HAT needs a separate regulator. The guide distinguishes the Pi 5 power arrangement from earlier boards. Budget for those components, storage, connectors, tools and the external treatment system; no current complete delivered cost was established. [7] [8]

Wiring instructions contain revision differences

Relay channel numbers must map to the actual board's physical pins. The guide's OWL driver-board example maps channels 0–3 to board pins 13, 15, 16, 18; its particular generic-HAT example instead uses 35, 33, 31, 29. Those examples are not universal Raspberry Pi GPIO mappings. Select the actual HAT and numbering convention before applying a configuration. [7]

The same guide names Fathomlock connectors in the parts table, EcoMate Aquarius in assembly, and a Bulgin connector in the later solenoid section. These references do not establish one interchangeable harness. Exact connector selection and pin correspondence remain unresolved in this review; the article is not a validated wiring recipe. [7]

Commission the camera and storage before closing the enclosure

The guide recommends software setup before inserting the tray because display and USB access become harder afterward. The installer route uses 64-bit Raspberry Pi OS and offers basic, standalone-dashboard and networked configurations. It changes system packages and startup services; none of those operations was executed here. Focus the camera at the intended working distance and inspect detection output before relying on actuation. [7] [9]

The getting-started page requires a USB drive. The inspected source adds useful detail: controller or dashboard configurations force recording infrastructure setup, and storage failures can leave the process alive while it signals an error. Consequently, an apparently running service alone does not prove that detection is ready. Check storage and application logs as well as service status. This source excerpt does not establish a universal USB requirement for every possible headless configuration. [8] [9] [12]

Detection performance depends on the task

The color route uses vegetation-color thresholds and contours to assign detections to output zones. It is useful for separating green vegetation from a contrasting background, but cannot by itself distinguish a green crop from a green weed. Camera focus, lighting, motion blur, target size and detection-to-actuation timing remain application-specific. No field accuracy, treatment efficacy or operating-speed claim was independently tested for this entry. [5] [10]

A separate green-on-green option exists in the current project, but it adds model and dependency requirements. Its weights, datasets, accelerator compatibility and licensing were not qualified here. They are not included in a claim that the base color detector is reproducible. [5] [13]

The main repository is MIT licensed, and the separate documentation repository explicitly adopts that license. Printable models are present in the archive; external boards, purchased components and optional learned-model dependencies need their own assessment. No enclosure was printed, wiring energized, software installed or treatment equipment operated during this review. [5] [6] [11] [13]

What we checked

Source research only; no installation, physical build or performance test performed. No affiliate links added.

Not yet established

  • Requirements: tools
  • A step-by-step setup summary

Updates

Upstream last updated
Apr 4, 2026 [2]
Latest release
Apr 4, 2026 [2]
Profile updated
Oct 25, 2026
Sources last checked
Oct 1, 2026
Next review
Oct 31, 2026
Reviewed revision
46296531dba1474887b1238c73f9cf83ac41c47e source branch; documentation 939f9bfabb62633a3d894219495b86793b8e9180 · combined hardware and software repository

Source files and licenses

Available project sources support research coverage; complete build and commercial qualification remain separate.

LayerAvailability / license
mechanical designavailable · MIT
Repository project license; external components and dependencies require their own terms. Models inventoried, not geometrically validated. [5] [6] [13]
electronicsnot found · License not fully assessed
Purchased Pi/camera/relay components and separate official driver-board sources are outside the matched artifact review. Not a closed-source finding. [7] [13]
firmwarenot found · License not fully assessed
Purchased Pi/camera/relay components and separate official driver-board sources are outside the matched artifact review. Not a closed-source finding. [7] [13]
softwareavailable · MIT
Repository project license; external components and dependencies require their own terms. Models inventoried, not geometrically validated. [5] [6] [13]
model weightsnot found · License not fully assessed
Optional learned-model route is not qualified here; matched model/data provenance and licenses remain unreviewed. Selected color-detection route does not require learned weights. [5] [13]
datasetsnot found · License not fully assessed
Optional learned-model route is not qualified here; matched model/data provenance and licenses remain unreviewed. Selected color-detection route does not require learned weights. [5] [13]
documentationavailable · MIT
Documentation README explicitly refers to main project MIT license. [6] [11]

References

Each source links to the exact revision we read. “Used for” shows where this page relies on it.

Correction guidance · Discussion is not enabled · Profile format 0.6.0-draft.1

Change history

2026-10-25 — Added an attributed upstream image with linked reuse terms; technical evidence dates unchanged.