OpenTEAM and AgStack: Farmer-Owned Data Platforms

The Consortium as Shared Infrastructure

Healthy ecosystems run on shared infrastructure. Mycorrhizal networks move phosphorus between species that never interact directly. The atmosphere cycles nitrogen through bacteria, weather, and vegetation without any single organism controlling the flow. OpenTEAM applies the same logic to agricultural data: a shared infrastructure layer where the farmer is the principal and the data moves through agreed protocols rather than proprietary silos.

OpenTEAM, the Open Technology Ecosystem for Agricultural Management, launched in 2019 with Wolfe's Neck Center for Agriculture and the Environment as its founding operational anchor. Wolfe's Neck runs a 626-acre certified organic and grass-fed beef operation in Freeport, Maine, and became the first site to pilot the consortium's data-sharing framework at commercial scale. By 2021 the enrolment base had reached approximately 250,000 acres across member farms; by 2024 it exceeded 1.5 million acres under farmer-owned data standards (OpenTEAM 2024 Impact Report). Partners include Stonyfield Organic, Clif Bar, the Rodale Institute, and the National Young Farmers Coalition, as well as research institutions and supply chain buyers who receive data access under terms the individual farmer approves.

The distinction from a conventional agricultural cooperative is structural. In most cooperative arrangements, the cooperative entity holds pooled member data and licences it to third parties or uses it for collective bargaining. OpenTEAM's governance documents are explicit: each farmer is the data principal. Research institutions and supply chain partners receive access only to the specific datasets each farmer permits, under terms governed by Creative Commons or custom bilateral agreements. The consortium coordinates interoperability; it does not aggregate data under its own ownership (OpenTEAM Governance Framework v1.2, 2023).

The AgStack Standards Layer

Every farm data platform has its own internal data model. A soil test result recorded in Climate FieldView exists in a format that cannot leave Climate FieldView without Bayer's permission. A planting record in John Deere Operations Center uses proprietary schema that John Deere controls. The farmer who generates both records owns neither in any portable sense. AgStack exists to make this problem technically solvable without requiring cooperation from the platforms that benefit from the lock-in.

The Linux Foundation launched AgStack in 2021 as an open-source, vendor-neutral project to build the infrastructure layer for agricultural data interoperability (Linux Foundation 2021 Project Announcement). The Foundation's governance model is critical to why this matters: no single corporation controls the specification. Changes require community consensus. The code is released under the Apache 2.0 licence, which means any operator, cooperative, or tool developer can implement the standard without paying a licence fee or seeking permission from a private entity. Contributors as of 2024 include Bayer Crop Science, IBM, Trimble, and a cohort of open-source developers, operating under Linux Foundation neutral-territory governance that prevents any single contributor from steering the standard for private advantage (AgStack GitHub repository, Linux Foundation 2024).

What AgStack actually provides: open API specifications that define how agricultural data objects, field boundaries, soil records, crop observations, equipment telemetry, and weather station readings should be structured and exchanged between different software systems. A tool built to the AgStack API spec can send data to, and receive data from, any other AgStack-compatible tool without a proprietary translation layer. This is the technical precondition for farmer data portability at scale.

OurSci and FarmOS: the Operator Toolkit

AgStack provides the plumbing specification. OurSci and FarmOS are the tools the operator actually picks up in the field. OurSci is an open-source survey and data collection platform designed specifically for regenerative agriculture metrics. Wolfe's Neck Center developed its core measurement protocols in partnership with OurSci, and the platform has been deployed in approximately 2,000 survey contexts globally as of 2024, from smallholder assessments in the US Northeast to pasture biomass evaluations in New Zealand and agroforestry audits in Central America (OurSci 2024 Platform Report). The survey forms cover soil health indicators, cover crop establishment density, species diversity assessments, and biodiversity proxy counts. All data is collected in formats compatible with the OpenTEAM data layer and with the AgStack API specifications.

FarmOS completes the stack. Deployed across an estimated 2,500 to 4,000 operational farm installations globally, FarmOS is a self-hosted farm record system built on the Drupal content management framework by developer Mike Stenta (FarmOS.org 2024). It manages field boundaries, planting records, livestock movement logs, grazing events, equipment passes, and compliance documentation. The API FarmOS exposes is compatible with the OpenTEAM submission layer: an operator running FarmOS can configure direct data submissions to OpenTEAM-affiliated research programmes without any additional software. Self-hosting FarmOS on a basic virtual private server costs 5 to 15 EUR per month from providers such as Hetzner, DigitalOcean, or Linode. Against commercial farm management subscriptions that run 400 to 2,500 EUR per year depending on scale and feature set, the full open stack, including FarmOS hosting, OurSci survey tooling, and Sentinel-2 NDVI access at zero marginal cost, operates for under 200 EUR per year for most small to mid-scale operations.

The combination is not a workaround. It is a coherent architecture. FarmOS holds the farm record. OurSci generates the measurement data. The OpenTEAM layer submits agreed datasets to research and supply chain partners. AgStack ensures that none of these components requires the other to be from the same vendor.

Data Sovereignty as Operational Principle

The data sovereignty argument lands in the terms of service the operator signs on day one of a platform relationship. Climate FieldView, owned by Bayer following the acquisition of Climate Corporation for 1.1 billion US dollars in 2013, aggregates field-level telemetry across its user base and monetises it through insurance risk models and commodity price signals. John Deere Operations Center had enrolled more than 150 million acres globally as of the Deere Annual Report 2023, with machine telemetry and yield data aggregated under terms that grant Deere broad licence to use operator data for product development and market analytics. Granular, the farm management platform owned by Corteva Agriscience, follows similar data-use terms. In each case the operator generates the data, the platform extracts the value, and the operator pays both the subscription fee and the price of reduced informational advantage in input markets and insurance negotiations.

The Data Sovereignty spoke develops the full rent-layer analysis: the data layer is the seventh layer of the agricultural rent stack, sitting above land, inputs, equipment, credit, market access, and agronomic knowledge. OpenTEAM is where the exit from that layer becomes operationally concrete. An operator who builds on FarmOS, OurSci, and the OpenTEAM data framework does not need to renegotiate platform terms to export their own records. They never handed over the terms in the first place.

Where the Consortium Goes

The OpenTEAM enrolment curve from 250,000 acres in 2021 to 1.5 million acres in 2024 represents a compound annual growth rate of approximately 82 percent (OpenTEAM 2024 Impact Report). At that rate the network would cover 10 million acres before 2027. The practical ceiling is the rate at which farmers can access onboarding support and the rate at which supply chain partners formalise data-sharing agreements that make enrolment worthwhile. Both constraints are loosening. Rodale Institute and the National Young Farmers Coalition have published onboarding guides; Stonyfield and Clif Bar have formalised preferred-supplier premium arrangements linked to OpenTEAM-verified regenerative practice data.

The regulatory environment is shifting alongside the adoption curve. The European Union Data Act, which came into force in January 2024, establishes rights for business-to-business data portability that benefit farmers operating on captured platforms. Under the Act, users of data-generating devices, including connected tractors, soil sensors, and yield monitors, have a right to access the data their equipment generates and to transfer it to third parties, including open platforms such as FarmOS. US state-level Farmer Data Protection bills are active in eight or more legislatures in the 2023 to 2025 cycle, with most bills following a model statute that establishes farmers as the owners of all data generated on their operations (American Farm Bureau Federation 2024 Policy Tracker). Captured platforms that currently operate under broad data-use licences will face legal and regulatory pressure to support portability; operators who have already built on open infrastructure will face no transition cost when that pressure converts into enforceable rights.

The AgStack Foundation's technical working group is extending its API specifications into soil carbon MRV data portability, with the aim of allowing verified carbon sequestration records to move between farm record systems, carbon registries, and supply chain buyers without platform lock-in. This is directly relevant to the economics of regenerative certification: if carbon credit verification data is portable, the operator can work with multiple registries and switch methodologies without rebuilding their data programme from scratch. The data infrastructure for regenerative agriculture and the infrastructure for climate-aligned supply chains are converging on the same open standards.

The choice is not data or no data. It is data to the operator or data to the aggregator. OpenTEAM, AgStack, and OurSci are the infrastructure that makes the operator side of that choice functional at scale. The farmer who generates the record is the principal. The record goes where the farmer permits. That is not an ideological position. It is a governance design, implemented in code, running on 1.5 million acres.