Drone surveying pesticide-free crop fields with disease segmentation overlays and sensor readings
← All case studies

Computer Vision · AgriTech IoT · Case study

Eight Orbit Urban Farms

AI Crop Disease Detection & Real-Time Sensor Alerts at 1,500-Acre Scale

CodeSquad built Eight Orbit a two-stage computer vision pipeline and a real-time threshold engine. Together they find crop disease, water stress and nutrient imbalance across 18 pesticide-free fields and alert farm managers before problems spread.

Services Computer Vision · Threshold & Alert Engines · IoT Data · Cloud Platform
1,500 acres18 active fields monitored by one platform
2,100+ sensorsSoil, weather and aerial telemetry, triaged automatically
5 monthsProduction MVP delivered in 10 Agile sprints
The challenge

No pesticides means no second chances.

Eight Orbit grows without pesticides, uses 95% less water than conventional farming and reaches up to 6x higher yields. When disease appears, it can’t spray the field. Its tools are isolation, irrigation and nutrition, and they work best while the problem is still small.

At 1,450–1,500 acres across 18 fields, manual scouting couldn’t keep up. By the time early blight or powdery mildew was visible to the eye, it had often already spread to neighboring plants.

The operation was designed to run 1,800–2,100 soil sensors, 22–45 weather sensors and 4–8 aerial crop health sensors. Without automated triage, a sudden moisture drop or a building nitrogen deficit was lost among thousands of normal readings.

Eight Orbit needed two things working together: visual diagnosis that names a disease and maps where it is, and real-time sensor triage that leaves managers with a short list of what to act on now.

Deep dive 1 · Computer vision

Two models: one names the disease, the other maps it.

Drone imagery flows S3 → AI → results → portal. A classifier identifies the disease, and a segmentation model draws its exact boundary in the field.

  1. Capture

    Drones carry MicaSense RedEdge multispectral and FLIR Vue TZ20 thermal cameras, alongside standard RGB crop imagery.

  2. Ingest

    Every image and video lands in AWS S3, traceable to its field, flight and capture time.

  3. Classify

    EfficientNetB5 reads high-resolution RGB images and labels specific diseases such as Blight and Powdery Mildew.

  4. Segment

    A UNet with a ResNet backbone processes multispectral and RGB imagery into pixel-level masks of the diseased areas.

  5. Surface

    Labels, confidence scores and masks appear in the portal, and new disease regions go straight to the alert engine.

InputSignal
MicaSense RedEdgeRed-edge & NIR vigor changes
FLIR Vue TZ20Canopy temperature / water stress
RGB imageryLesion & coating patterns
ModelAnswers
EfficientNetB5What disease is it?
UNet + ResNetExactly where is it?
Classification without segmentation can’t say which rows to treat. Segmentation without classification can’t say what the problem is.
Deep dive 2 · Threshold & alert engine

Thousands of readings, cut down to a few clear actions.

Telemetry flows Field API → time-series → stats, then thresholds → dashboard notifications. Django APIs and async background workers do the evaluation.

  1. Ingest

    Soil and weather sensors from several vendors report through one Field API, normalized into a single schema.

  2. Aggregate

    Readings are stored as time-series and rolled up into PostgreSQL statistics: rolling averages, rates of change and per-field trends.

  3. Forecast

    LightGBM, Random Forest and GBM models forecast soil moisture and rainfall from the sensor and weather history.

  4. Evaluate

    Background workers check live stats, CV disease masks and forecasts against strict agronomy thresholds.

  5. Alert

    When a threshold is crossed, an alert naming the field, the trigger and the data behind it is pushed to the manager’s dashboard.

HardwareMeasures
Decagon EC-5Soil moisture
Yara N-TesterNPK nutrients
PT1000Soil temperature
Davis Vantage Pro2Weather station
Campbell Sci. ARG100Rainfall
TriggerAlert
Sudden moisture dropIrrigate (no rain forecast)
NPK imbalanceTargeted fertilization
Early blight maskInspect & isolate
Forecasts let the engine tell a dry spell that needs water now from one that incoming rain will fix.
Platform & production hardening

Built to keep working in the field.

Dashboard

React + TypeScript: Soil Health Score (92/100), NPK bars, yield (28 t/ha), precipitation vs. moisture

Data

Django APIs over PostgreSQL time-series stats, with raw imagery and model artifacts in AWS S3

Compute

Containerized services on AWS ECS/EC2, so APIs, workers and inference scale independently

Monitor

CloudWatch shows ingestion lag, worker backlogs and failed jobs before they affect alerts

Secure

TLS encryption in transit and role-based access for field staff, agronomists and management

Ship

Full CI/CD across Dev, Staging and Production, so model and threshold changes are tested first

Execution roadmap

Five milestones, ten sprints, five months.

Each milestone delivered a working, demonstrable piece of the system.

MilestoneSprintsDelivered
M1 · Core System1–2Django, PostgreSQL, Field API, time-series pipeline, AWS & CI/CD
M2 · AI Disease Detection & Forecasting3–4EfficientNetB5, UNet + ResNet, LightGBM / RF / GBM
M3 · Drone Integration5–6MicaSense & FLIR imagery via S3 into inference
M4 · Complete Dashboard7–8React + TypeScript portal with live alerts
M5 · Testing & Pilot Deployment9–10End-to-end testing, threshold tuning, pilot rollout
Impact

Proactive irrigation. Moisture alerts combined with rainfall forecasts tell managers which fields need water and which incoming rain will cover. Eight Orbit’s water-efficient model depends on that precision.

Targeted fertilization. NPK monitoring flags imbalances field by field, so nutrients go only where the data shows a deficit.

Zero-pesticide containment. Named diseases with pixel-level boundaries let crews isolate early blight and powdery mildew while outbreaks are still small enough to contain without chemicals.

One operational picture. Drone imagery, soil telemetry, weather and forecasts appear on one dashboard, with the system deciding what needs attention first.

FAQ

Common questions.

Does this only work for agriculture?

No. The same pattern applies anywhere images and sensor data arrive faster than people can review them: classify the defect, segment where it is, then check it against thresholds and alert the right person. That includes manufacturing inspection and infrastructure monitoring.

Can it work with the sensors and cameras we already have?

Yes. The Field API normalizes readings from different vendors into one schema, so new hardware is added as an adapter rather than a rebuild.

Who sets the alert thresholds?

Thresholds are agronomy rules defined with your domain experts and tuned during the pilot. They’re configured and promoted through Dev, Staging and Production like any other change.

How long did it take?

A production-hardened MVP in 5 months, delivered in 10 Agile sprints across 5 milestones.

Images and sensor data piling up faster than your team can review them?

Let’s turn them into decisions your team acts on.

Book a Free Call →