Drones and Remote Sensing in Agriculture Training Course

Participants can structure farm spatial data for accurate imagery analysis and reporting.

• Geographic and projected coordinate reference systems selected for farm mapping
• Datum transformations checked before combining GNSS, imagery, and field boundaries
• UTM zones, units, and map scale interpreted for local operations
• Raster cells, vector features, and attribute tables structured for analysis
• Georeferencing residuals assessed using root mean square error
• Metadata recorded using ISO 19115 elements and acquisition parameters

Participants can prepare farm datasets that support traceable spatial analysis.

• Field boundaries digitised from imagery and verified against farm records
• Farm data model designed for blocks, seasons, varieties, and treatments
• Attribute domains standardised for crop codes, units, and missing values
• GNSS observations imported with timestamps, accuracy fields, and identifiers
• Quality-control rules flag duplicate geometries, gaps, overlaps, and invalid attributes
• Versioned project folders preserve source files, derivatives, and processing notes

Participants can select drone platforms and payloads against field and data requirements.

• Multirotor, fixed-wing, and VTOL platforms compared against field requirements
• Payload capacity, endurance, wind limits, and launch constraints calculated
• Sensor-platform pairing determined from required ground sample distance
• Battery logistics planned using mission duration, reserves, and charging cycles
• Airspace restrictions, site hazards, and privacy obligations documented before flight
• Contracted-service specifications define coverage, accuracy, turnaround, and deliverables

Participants can match sensor characteristics to crop-monitoring questions and accuracy needs.

• RGB, multispectral, thermal, and LiDAR sensors matched to crop questions
• Spectral bands related to chlorophyll, canopy structure, water, and soil
• Radiometric resolution and dynamic range assessed for subtle stress detection
• Lens focal length linked to altitude, footprint, and ground sampling distance
• Calibration panels and downwelling sensors support reflectance consistency
• Thermal measurements corrected for emissivity, timing, and atmospheric conditions

Participants can prepare flight plans that meet coverage, accuracy and safety requirements.

• Mission polygons, buffers, altitude, overlap, and sidelap configured for orthomosaics
• Ground control points distributed across terrain and image-block edges
• RTK and PPK workflows selected for positional accuracy requirements
• Pre-flight checklists cover weather, batteries, firmware, airspace, and observers
• Flight lines oriented to wind, sun angle, and crop geometry
• Contingency procedures define lost-link, flyaway, battery, and weather responses

Participants can apply operational controls to reduce flight, privacy and data-handling risks.

• Operational risk assessment identifies people, property, obstacles, and failure modes
• Permission pathways checked against local aviation authority and landowner requirements
• Visual line-of-sight, observer roles, and exclusion zones established
• Incident reporting, maintenance logs, and pilot records maintained for accountability
• Privacy-by-design controls govern imagery of workers, homes, and neighbouring land
• Data retention and access rules protect commercially sensitive farm information

Participants can collect field evidence that supports reliable image interpretation and validation.

• Field acquisition log captures sensor settings, light, wind, and anomalies
• Manual exposure and shutter speed set to limit motion blur
• Radiometric calibration images collected before and after each mission
• Ground-truth observations sampled alongside imagery using stratified field locations
• Phenological stage recorded using BBCH growth scale observations
• Repeat-flight protocol fixes altitude, timing, overlap, and sensor configuration

Participants can process drone image sets into documented spatial products for agricultural analysis.

• Image sets screened for blur, exposure variation, gaps, and duplicates
• Structure-from-motion processing aligns photographs into sparse point clouds
• Bundle adjustment minimises reprojection error using tie points and controls
• Orthomosaics generated with appropriate projection, resolution, and radiometric settings
• Digital surface models and canopy height models derived from point clouds
• Accuracy report separates checkpoints, control points, horizontal, and vertical error

Participants can prepare imagery and supporting spatial layers for field-scale analysis.

• Raster clipping, mosaicking, resampling, and reprojection prepared for field analysis
• Band stacking and layer symbology reveal patterns without misleading colour ramps
• Zonal statistics summarise vegetation indices within management zones
• Raster calculator expressions combine bands, masks, thresholds, and ancillary layers
• Terrain derivatives identify slope, aspect, flow paths, and erosion risk
• Cloud, shadow, and non-crop masks remove invalid pixels

Participants can select, calculate and interpret vegetation indices for crop-monitoring decisions.

• NDVI, NDRE, GNDVI, and SAVI selected for crop-stage questions
• Index equations calculated from reflectance bands with documented sensor wavelengths
• Histogram inspection detects saturation, skew, and unsuitable classification thresholds
• Relative vigour zones separated from absolute agronomic diagnosis
• Time-series index profiles compare treatment strips and seasonal trajectories
• Index limitations explained for dense canopies, soil backgrounds, and illumination

Participants can classify imagery and evaluate the reliability of mapped crop features.

• Unsupervised clustering creates preliminary zones from multispectral image features
• Supervised classification trained using labelled samples and balanced classes
• Random Forest parameters tuned using validation data and feature importance
• Confusion matrices report producer accuracy, user accuracy, and kappa
• Object-based image analysis groups pixels using scale, shape, and texture
• Class maps reviewed against ground observations before operational use

Participants can convert mapped anomalies into prioritised field-inspection and treatment decisions.

• Scouting routes prioritised from anomaly polygons, severity ranks, and access
• Pest, disease, nutrient, and irrigation hypotheses tested with field evidence
• Canopy temperature differences interpreted alongside weather and soil moisture
• Plant counts and gap detection estimated from high-resolution RGB imagery
• Early-season emergence maps separated from shadows, residue, and wheel tracks
• Decision thresholds linked to agronomist inspection and treatment protocols

Participants can use remote-sensing evidence to investigate soil variation and water-management issues.

• Drainage patterns mapped using elevation models, wetness indices, and imagery
• Thermal and multispectral evidence compared for water-stress diagnosis
• Soil sampling zones designed from stable spatial patterns and terrain
• Electrical conductivity and yield maps integrated with remotely sensed layers
• Variable-rate irrigation zones defined with agronomic constraints and system capacity
• Prescription assumptions documented before recommending field-specific input-rate changes

Participants can integrate yield and remote-sensing data to define defensible management zones.

• Yield monitor data cleaned for delays, overlap, moisture, and outliers
• Harvest maps calibrated against scale tickets and crop-moisture records
• Multi-year yield stability classified using mean yield and variance
• Management zones delineated from yield, soil, terrain, and imagery evidence
• Variable-rate prescriptions exported with units, equipment constraints, and audit fields
• On-farm trials designed with strips, replication, and spatial controls

Participants can assess whether drone-derived outputs are fit for their intended decisions.

• Reference samples designed using random, stratified, and targeted sampling schemes
• Sample size considered against class prevalence and required confidence
• RMSE, MAE, bias, and R-squared interpreted for continuous estimates
• Spatial autocorrelation checked before treating pixels as independent observations
• Uncertainty maps communicate confidence alongside predicted crop-condition surfaces
• Validation findings translated into collection changes and decision limits

Participants can produce clear maps and reports for agronomic and operational audiences.

• Cartographic hierarchy directs attention to decisions, anomalies, and recommended actions
• Map layouts include scale, north arrow, CRS, date, and source
• Class breaks selected using quantiles, natural breaks, and agronomic thresholds
• Field reports distinguish observations, interpretations, uncertainty, and recommendations
• Web-map layers prepared with clear names, filters, and permission settings
• Briefings frame financial, operational, and agronomic implications for managers

Participants can design operating arrangements for sustained agricultural drone data collection and use.

• Standard operating procedures define roles, handoffs, checks, and acceptance criteria
• Seasonal acquisition calendar aligns flights with crop stages and decisions
• Equipment maintenance schedule tracks batteries, sensors, firmware, and airframes
• Data governance assigns ownership, retention, access, and quality responsibilities
• Service-level measures track turnaround, coverage, accuracy, and usable outputs
• Business case compares in-house operations, contractors, and hybrid delivery

Participants can scope and plan a complete remote-sensing project before final production and review.

• Use case scoped with stakeholder question, crop, field, and decision owner
• Acquisition plan specifies platform, sensor, timing, controls, and safety measures
• Processing workflow selects photogrammetry settings, indices, and quality checks
• Analysis plan defines zones, field validation, and decision thresholds
• Report structure assigns maps, findings, uncertainty statements, and recommendations
• Peer review criteria assess traceability, agronomic logic, and usability

Participants can build, review and present a complete drone remote-sensing workflow and decision report for an agricultural field case.

• Project brief converted into a defensible drone remote-sensing workflow
• Flight and safety plan checked against operational and data-quality requirements
• Imagery products processed, documented, and assessed using stated accuracy criteria
• Management zones interpreted with ground evidence and uncertainty limits
• Decision report presents maps, recommendations, assumptions, and implementation steps
• Technical review addresses stakeholder questions, risks, and next-season improvements