Open-Source Agricultural Robots: Platforms Operators Own

FarmBot: CNC Precision Under an Open Licence

FarmBot is a Cartesian-gantry farming robot that moves across a raised-bed growing area on aluminium extrusion rails, executing pre-programmed planting, watering, and weeding sequences with sub-centimetre positioning accuracy. Rory Aronson published the initial design in 2013 during an agricultural systems degree at California Polytechnic State University, and the hardware and software have been iterated in public ever since. Hardware designs are released under Creative Commons licence terms, and FarmBot OS, the Raspberry Pi-based control software, is MIT-licensed. (Source: farmbot.io/open-source, 2024.) There is no proprietary firmware that requires specialist access and no dealer standing between the operator and their machine.

The Genesis XL model covers 1.5 metres by 18 metres and retails at USD 4,495 as a complete kit, including gantry, tool head, seeder, watering nozzle, and weeding implement. The Express XL, designed for smaller intensive operations, covers 1.5 metres by 3 metres at USD 1,695. (Source: farmbot.io, 2024.) Both ship with full assembly documentation. The design file is also the repair manual: every component is documented to part number, function, and tolerances, so the operator who builds the machine also holds the information needed to fix it at any future point.

The naturalist argument for CNC-level precision is not primarily about speed. A seed placed at exactly the correct inter-plant spacing eliminates competition between crop seedlings that undermines germination evenness in broadcast or loosely hand-sown beds. Per-position planting at 1-2 cm accuracy produces more uniform canopy closure, which in turn reduces the light available to weed seedlings at the critical white-thread establishment stage. The engineer notes that the architecture, a Cartesian gantry with swappable tool heads, is the same principle used in industrial CNC milling: the axis logic that positions a spindle over aluminium can place a tomato seed over compost at comparable precision. The operator argument closes the loop: a FarmBot Express XL running overnight plants and waters a 4.5 square metre bed in the time a grower would spend on other tasks, at a capital cost that amortises over a decade without software subscriptions.

Small Robot Company: Per-Plant Field Robots

Small Robot Company, founded in 2017 by Ben Scott-Browning and Sam Watson Jones in the United Kingdom, designed a trio of small autonomous field robots for row-crop applications: Tom for per-plant precision seeding, Dick for per-plant targeted spraying, and Harry for autonomous weeding using AI vision. The conceptual contribution is the per-plant frame itself. Rather than treating a wheat or vegetable field as a uniform input target requiring blanket application, each robot addresses every plant position individually. Per-plant spraying reduces total herbicide and fertiliser volume by more than 90 per cent compared with conventional blanket application (Small Robot Company technical documentation, 2022). The naturalist implication is direct: 90 per cent less active herbicide ingredient reaching the soil means 90 per cent less disruption to the microbial and fungal communities that process organic matter in the rhizosphere.

The company raised GBP 14.5 million in Series B funding in 2021 to support prototype field trials across wheat and vegetable operations in England. (Source: Companies House; SRC investor announcement, 2021.) In 2024 the company entered administration, with assets subsequently acquired. (Source: BBC News, 2024; The Grocer, 2024.) The administration is instructive rather than disqualifying. Per-plant precision mechanics in wheat and vegetable crops is now a demonstrated and documented capability, replicated by Ecorobotix's ARA system and in development at several other platforms. The argument does not depend on any single company surviving.

There is, however, a structural distinction that operators should note before drawing broader lessons from Small Robot Company's model. Their Farming-as-a-Service subscription model kept the hardware on the company's balance sheet. Farmers used the service; they did not own the robots. When the company entered administration, that service access ended. Operators who had built or purchased open-hardware equivalents retained full tool functionality on the same day. The architecture that persists through company failure is the operator-owned one. This is the same tension examined at greater depth in Equipment Sovereignty and the ECU Fence, which addresses the political economy of hardware access rights.

The Oggun Tractor: Open Design for Global Agriculture

The Oggun is a Paraguay-designed open-source tractor, developed by Horacio Clemente through Cleber LLC from 2011 onwards, to give smallholder and cooperative farmers access to a repairable, affordable alternative to proprietary imported machinery. The design targets 17-18 horsepower diesel output, sufficient for basic tillage, bed preparation, and transport tasks on operations up to approximately 20 hectares. Hardware design files are publicly available. The machine is designed to be fabricated from generic metal components available through standard hardware supply chains in contexts with limited access to specialised agricultural dealers. Estimated build cost from the published design documentation is approximately USD 3,000-5,000 depending on local materials pricing and fabrication labour. (Source: Oggun Project documentation, 2011-2023; Clemente design filings; Cleber LLC public build records.)

The engineering principle that distinguishes the Oggun from a commercial proprietary tractor is not horsepower or travel speed. It is the relationship between the machine and the information required to repair it. Every component in the Oggun design exists in publicly available documentation with its function, tolerances, and generic equivalent specifications. When a tractor built to this design fails, the operator who assembled it holds all the information needed to fix it, using tools and parts available from general metalworking suppliers. When a John Deere 8R fails, the electronic control unit may require dealer authorisation before any diagnostic or repair step can proceed. That distinction is the central argument developed in Equipment Sovereignty and the ECU Fence. The Oggun is not faster or more capable than a modern commercial tractor. It is owned, in the fullest sense of the word.

Smallholder cooperative systems with restricted access to proprietary parts networks and authorised dealer chains benefit most directly from this architecture. The Oggun's licensed export to Cuba through Cleber LLC made this deployment context explicit: a tractor whose entire repair pathway is accessible to any cooperative with basic metalworking capacity is categorically different from one whose diagnostic software is held under licence by the original manufacturer. The naturalist register matters here too. A tool ecosystem that depends on a single external supplier for access and repair is fragile in proportion to that dependency. A locally repairable, locally documented tool ecosystem is as resilient as the community using it.

Solectrac: Electric Compact Traction on Operator Terms

Solectrac is a California-based company founded by Steve Heckeroth to develop electric compact tractors for small-scale and specialty crop operations. The e25G, rated at 25 horsepower equivalent, carries a manufacturer's suggested retail price of USD 46,995 (Solectrac/Ideanomics, 2024). The eNarrow, designed for the narrower corridors of orchard and vineyard operations, is available at USD 24,995 (Solectrac, 2024). Ideanomics acquired the company in 2021 to expand its electric vehicle portfolio into the agricultural sector.

Solectrac is not an open-source project; the design is proprietary. The reason it belongs alongside FarmBot and Oggun in this analysis is that electric tractor architecture is, by design, more operator-maintainable than a diesel engine paired with a proprietary electronic control unit. The drivetrain is a 48V DC system, serviceable by any competent electrician with knowledge of DC motor and battery maintenance. The battery management system does not require manufacturer authorisation to diagnose. The removal of the diesel fuel system eliminates an entire category of wear-part replacement and the associated dealer dependency for fuel system components. An operator on a regenerative farm running on-farm solar generation has a further integration argument: electricity from owned generation infrastructure powers owned machinery, with no fossil fuel invoice in either direction.

The USDA Natural Resources Conservation Service ran electric compact tractor evaluation programmes in partnership with Solectrac from 2021, documenting performance across mixed vegetable, vineyard, and small-grain contexts. The evaluation data confirms that 25 HP equivalent is sufficient for the tillage and transport tasks that account for approximately 70 per cent of compact tractor use hours on operations under 40 hectares. (Source: USDA NRCS electric tractor evaluation programme, 2022-2023.) The electric tractor is not the right tool for every task on every farm. For the majority of tasks on small-to-mid-scale regenerative operations, the electrification argument and the operator-maintainability argument point in the same direction.

The Farm-Scale Open-Source Movement: What Ownership Means

The broader movement within which FarmBot and Oggun operate has an institutional form. Farm Hack (farmhack.org) is an open-source farm tool design community that has documented more than 300 tools since its founding in 2012, ranging from simple soil blockers and seeding jigs to cover-crop rollers, cultivators, and small processing equipment. Farm Hack designs are released under Creative Commons Attribution licences, meaning any farmer can build, modify, and share modifications without restriction. (Source: farmhack.org, 2024.) The network operates in partnership with Cornell University Cooperative Extension and the Natural Resources Conservation Service.

Open Source Ecology, founded by Marcin Jakubowski, pursues a more systematic version of the same argument. The Global Village Construction Set (GVCS) is a published set of 50 industrial machines, including the LifeTrac open-source tractor, designed with interlocking modular components that any community with basic fabrication equipment can build. The LifeTrac design is released under CC BY-SA 4.0: any operator who fabricates one and modifies the design must release those modifications under the same terms. (Source: Open Source Ecology documentation; LifeTrac v18 design file release, 2021.) The licence is the mechanism by which the commons grows rather than being captured back into proprietary form.

The operator economics of the open-source hardware path resolve around one structural difference from the proprietary path: depreciation stops but functionality does not. A FarmBot Express assembled for USD 1,695 in 2024 will still perform its original function in 2034 if maintained, because the design files, documented replacement parts, and community knowledge base are publicly accessible. The proprietary equivalent at two to four times the capital cost may require dealer network support that is discontinued if the original manufacturer exits the market or discontinues the product line. Farm robotics companies enter administration. Open hardware licences do not.

The relationship between open hardware and the right-to-repair legal environment is complementary rather than redundant. Right-to-repair legislation at the federal level (including the FTC's 2024 enforcement framework) and the cascade of state-level repair bills across Nebraska, Colorado, New York, and Massachusetts forces proprietary manufacturers to provide diagnostic access and parts availability that was previously withheld. Open hardware avoids the need for that legislation entirely, because the design was never locked in the first place. Operators who choose the open path are not waiting for regulation to restore access. The machinery they own already answers to them. For the detailed account of the legal and economic dimensions of that access gap, see Equipment Sovereignty and the ECU Fence.