2602 Woodlawn Holdings — McKinney, Texas

01 · The Land

Seventeen acres on the East Fork Trinity, minutes from McKinney.

The parcel at 2602 Woodlawn Road sits at the hinge between agricultural Collin County and the expanding footprint of McKinney. Pasture and row crop to the north, the East Fork Trinity River corridor to the south, and a natural oxbow pond folded into the middle of the property.

Water, shade, and a working farm road — the ingredients of a living farm are already here.

Of 17.38 total acres, 5.5 sit on higher ground suitable for production. The remaining acreage stays intact: mature canopy, riparian edge, and the pond itself serve as the biological buffer for the whole system. Nothing about the land is asked to change — the farm is asked to fit it.

Address

2602 Woodlawn Road (CR 294), McKinney TX

Parcel

PID 2802080 · Collin County

Acreage

17.38 total · 5.5 usable production zone

Water

~2.5 acre natural oxbow · East Fork Trinity tributary

Access

684 LF farm road · Woodlawn frontage

Context

Rural edge · 20 min from downtown McKinney

The oxbow · biological anchor

Site overview · west approach

Canopy cover · riparian edge

Woodlawn Rd frontage · 684 LF

02 · The Vision

Farm the way the land already works.

Not a factory on top of the field. A farm shaped by the pond, the canopy, and the seasonal rhythms that have been running on this parcel for longer than any deed. The vision is a modern VAC — Vườn · Ao · Chuồng, garden · pond · coop — built as one living, closed-loop organism.

Tenet 01

The pond leads the design.

The oxbow is not amenity — it is the biological reactor everything else plugs into. Built components sit downstream of its microbiology. Site decisions begin with water and work outward.

Tenet 02

Instrumented, not industrialized.

Sensors and inference replace constant human presence. The land keeps its rhythms. The operator gains a quiet, continuous view into them — and the ability to ask the system, in plain language, what is going on.

Tenet 03

A closed, learning loop.

Fish feed the greens. Greens feed the table. Birds feed the soil. Every intervention is logged; every outcome feeds back. What the land grows, the land keeps — and what the system learns, the system remembers.

03 · The Water

A 2.5-acre oxbow, quietly doing three jobs.

The oxbow is a natural meander cut off from the East Fork Trinity. Fed by the riparian water table, shaded by old canopy, and seeded with decades of native microbiology — the kind of water you cannot build.

Biological seed.

Pond water seeds the aquaculture biofilter. Native microorganisms accelerate system cycling by weeks and displace a meaningful share of the chemical mineral inputs otherwise required.

Buffer reservoir.

When tank chemistry drifts, partial water exchange with the pond stabilizes ammonia and pH without discarding treated water — trading restart risk for quiet self-correction.

Nutrient endpoint.

Composted manure tea returns to the pond margins — nitrogen back into the native water body. The loop closes at the place it started. Zero synthetic discharge off the parcel.

Oxbow · looking south from farm road

04 · The VAC

Garden. Pond. Coop.
One organism, three parts.

Vườn · Ao · Chuồng — the Vietnamese integrated-farming tradition rebuilt with modern controlled-environment agriculture. Three production components share water, nutrients, and waste as one closed loop. This is the fundamental principle of the farm — every sensor, every model, every delivery lives downstream of it.

Garden · Vườn

Controlled-Environment Greenhouse

A single-span polycarbonate greenhouse with deep-water culture grow channels, LED supplemental lighting, and HVAC venting. Nutrient-rich water from the aquaculture tanks circulates directly through the channels — no synthetic fertilizer, no seasonal break, no waste stream.

Pond · Ao

Recirculating Aquaculture + Natural Oxbow

Indoor recirculating tanks paired with the natural oxbow as a biological and hydrological buffer. The pond seeds the biofilter; the tanks concentrate the nutrients the greenhouse needs to live. The pond anchors the loop.

Coop · Chuồng

Laying Flock · Compost Loop

A ventilated poultry house sited against the pond edge. Manure is channel-collected, composted with crop waste, and returned to grow media and pond margins — the detail that closes the nutrient loop back to the place it started.

05 · The Technology Vision

Aquaponics 4.0 — the intelligence layer.

Replace the full-time on-site biologist with a dense IoT sensor network feeding a real-time AI inference layer. Sensors read every thirty seconds. Edge devices apply local control rules instantly. Cloud models detect drift hours before critical thresholds. LLM agents translate sensor data into natural-language root-cause hypotheses. A farm queried the way a network operations center queries its telecom stack.

→

Biology becomes queryable.

Water chemistry, fish behavior, and canopy growth arrive as live data streams. The land keeps its rhythms — the operator gains a window into them.

→

Abstraction over the field.

The knowledge base is the farm's operating system — abstracting biological complexity the way a network OS abstracts hardware from an application.

→

Right-sized by design.

The architecture is modular. Every layer is buildable at pilot footprint and extensible from there. Small, instrumented, and alive.

Five-Layer Architecture

Physical

Sensor mesh across water, air, and crop.

Atlas Scientific pH, dissolved oxygen, electrical conductivity, and ORP probes. Ultrasonic water level. IP cameras on fish behavior and canopy imaging. Temp and humidity. Dual-redundancy on every life-critical channel. ESP32 nodes transmit via MQTT with battery backup at each node.

pH ±0.01DOECORPUltrasonic levelIP camerasESP32 · MQTT

Edge

Local control that never waits on the cloud.

Raspberry Pi 5 running Mosquitto + Python actuator scripts. Sub-second response to critical events, hardware-capped dosing, 72-hour local buffer. The system degrades gracefully; the land does not.

Raspberry Pi 5MosquittoRule engineHardware safety caps72h buffer

Cloud

Every reading, tagged, queryable, permanent.

AWS IoT Core → InfluxDB Cloud → Grafana. Raw 30-second data kept at fidelity; rolled into aggregates; daily summaries retained indefinitely. Every reading carries tank, sensor, and crop-batch identity. A system whose history can be asked questions.

AWS IoT CoreInfluxDBGrafanaTagged time-series

ML Models

Inference that sees drift before humans do.

A random-forest fish-health classifier on rolling water chemistry. An LSTM anomaly detector that flags systematic deviation hours before threshold. YOLOv8 canopy vision estimating days-to-harvest. Feed optimization tied to fish activity and oxygen draw. Every model retrained on this farm's own data.

Random forestLSTMYOLOv8Weekly retrain

LLM Agent

“Why is tank 2 showing stress this morning?”

Claude API with function calling, wired into the sensor pipeline. The operator asks in plain language; the system retrieves 48 hours of sensor history plus the most relevant passages from the RAG knowledge base, and returns a grounded root-cause hypothesis with a recommended action. A daily operations summary at dawn.

Claude APIFunction callingRAG · ChromaDBDaily ops brief

06 · The Knowledge Base

A farm that learns what works on this land.

The knowledge base compounds in three layers. Live operational data grounds it in this parcel's own behavior. Curated science and regulation keep it honest. A structured log of every intervention and its outcome turns the system from static reference into something that actively improves.

Live operational data.

Every sensor reading, every anomaly, every corrective action, every harvest outcome — permanently logged and queryable. A dataset that can only come from this facility, training every subsequent model on ground truth from this specific pond and this specific canopy.

Curated science, via RAG.

Peer-reviewed aquaponics literature, Atlas Scientific application notes, USDA organic guidelines, FDA Produce Safety documentation, FEMA floodplain guidance — chunked, embedded, retrieved alongside live state. Responses grounded in both real-time system and authoritative science.

Decision outcomes.

A structured record of every intervention and what followed. Over time, a ground-truth dataset of what actually works on this parcel — which alerts were false positives, which varieties underperformed the model, which corrections held. The base stops being static and starts learning.

07 · Research Foundation

Not a first attempt. A well-documented one.

Integrated VAC (vegetable–aquaculture–chicken) farming and pond-based polyculture are the subject of a working peer-reviewed literature, and commercial aquaponics is now a three-decade industry with operators producing food year-round in climates far harder than North Texas. Woodlawn's architecture rests on that base — not on novelty.

Evidence · Working Commercial Systems

Nelson & Pade, Inc. — Montello, Wisconsin

Founded in 1984 and relocated to Wisconsin in the mid-2000s, Nelson & Pade operate a 14,000-square-foot controlled-environment aquaponics greenhouse that produces 1,000 heads of lettuce and 100 pounds of fish every week, year-round, alongside tomatoes, herbs, kohlrabi, and radishes. Their facility also houses the University of Wisconsin–Stevens Point Aquaponics Innovation Center, a formal academic–industry research partnership in operation since 2014. Their patented Clear Flow Aquaponic Systems are deployed in roughly 35–40 countries and have been recognized three times as Sustainable Product of the Year by the Wisconsin Sustainable Business Council.

14,000

sq ft greenhouse
continuous production

1,000 / 100

lettuce heads &
lb of fish · weekly

6×

yield per acre vs.
Gila Valley lettuce

⅙

of the water,
same acre basis

35–40

countries with
deployed systems

40+ yr

operating history
since 1984

The relevance is simple: a closed-loop aquaponic facility producing a thousand retail-grade lettuce heads a week is not a thought experiment. It has been running, under independent academic observation, for over a decade. Woodlawn's VAC is a smaller, outdoor, pond-coupled variant of the same fundamental principle.

Literature · Peer-Reviewed

VAC · Integrated Farming

Transformation to integrated VAC (Vegetable–Aquaculture–Chicken) farming system

Channa, A. A. & Tariq, M.

2022 · International Journal of Agriculture & Environmental Sciences

Frames VAC as an integrated food-production system where fishpond, vegetable plots, and poultry cycle nutrients through each other — the same loop logic Woodlawn inherits.

Polyculture · Smallholder

Pond-based integrated farming: a sustainable approach to low-input food production

Anila, S. & Daramola, J. A.

Nigerian Journal of Fisheries & Aquaculture · polyculture & integrated systems literature

Documents the economics and resource efficiency of pond-centered integrated farms at smallholder scale — the operational precedent for a single-pond, single-parcel loop like Woodlawn's.

Aquaponics · Systems Review

Aquaponics as a sustainable food production system: a review

Lama, et al.

Reviewed in the aquaponics sustainability literature · water reuse, nutrient cycling, fish–plant ratios

Consolidates the evidence for aquaponics as a sustainable production mode: closed-loop water use, stacked yields, and the component ratios that commercial operators like Nelson & Pade have since validated at scale.

Additional sources in the knowledge base: Atlas Scientific probe application notes, USDA National Organic Program standards, FDA Produce Safety Rule guidance, FEMA floodplain management documentation. Every citation is embedded into the RAG layer so operational answers can be grounded in the source.

08 · A New Way To Produce Food

Instrumented land. Open-loop learning. Quiet operation.

From labor to leverage.

Edge control, cloud inference, and language interfaces compress the operating team while expanding what a single farm can safely know about itself.

From recipes to observation.

Practices stop being inherited and start being measured. Every model is retrained on this farm, this pond, this climate — not someone else's benchmark.

From output to understanding.

The point of the intelligence layer is not more volume. It is knowing, in language, why the system is doing what it is doing — and when to leave it alone.

From farm to platform.

The sensor mesh, the knowledge base, the LLM interface — these are the transferable assets. Woodlawn is the first instance of an architecture meant to travel.

A pond,a greenhouse,a closed loop.

The land, from above.