KB

AgTech - Precision Agriculture: Lecture 2

RFID in Livestock

  • Purpose: identification, traceability, data integration with herd management and NLIS (National Livestock Identification System).
  • Tag types: Active or Passive (NLIS is passive).
  • Timeline: RFID introduced in livestock around 1999; compulsory tagging for cattle; also used for sheep.
  • Uses: track movements property-to-property, assist management decisions, treatments, and performance data.
  • How it works (passive): tag, reader, antenna; energy from antenna wakes tag; tag transmits data to reader.

Walk-over-Weigh (WOW)

  • Concept: weight data captured as animals walk over a weigh platform with load cells; RFID tag read by panel reader during passage.
  • Data: liveweight logged and linked to tag number; typically located at water points or attractants.
  • Hardware overview (box on page): single-board computer, battery, 4G LTE modem, charger controller, RFID panel reader, RFID antenna, weighing platform, load bars.
  • Benefits: daily individual weight monitoring, rapid management interventions, reduced weight loss, identify poor performers, lower labour for mustering.
  • Challenges: learning curve (e.g., cows ~10 days), multiple animals on platform, getting all sheep to use it, weight accuracy, data transfer.

OptiWeigh

  • Description: portable, movable in-paddock weighing system.
  • Benefits: voluntary use, minimal labour, reduced site setup; in-paddock weights.
  • Limitations: only part of mob weighed; need training to use a portion of mob; requires attractant.

On-animal Sensors

  • Components: GPS/GNSS for location; accelerometer for behaviour; 3G/4G connectivity for data transmission.
  • Concept: continuous monitoring of location and activity to infer health, welfare, and production signals.
  • Example: Sheep Health Monitor shows real-time alerts (e.g., sick sheep) via mobile messages.
  • Advantages: 24/7 monitoring, early issue detection, behavioural insights.
  • Challenges: device robustness, power consumption, data processing, and connectivity.

Virtual Fencing (VF)

  • Principle: audible warning + electrical stimulus when an animal crosses a virtual boundary.
  • Zones: Permitted Zone vs Exclusion Zone (EZ).
  • Potential benefits: environmental protection (riparian areas), precise pasture management, labour reductions, monitoring of movements.
  • Challenges: animal welfare considerations, learning for all animals, longevity of sensors, form factor, public perception.

Plant Sensing: Remote vs Proximal

  • Why: substantial variation in crop yield; sensing helps manage variability for future seasons.
  • Terminology:
    • Proximal = close to plants
    • Remote = from a distance
    • Active = uses its own light source
    • Passive = relies on external light sources
  • Sensing categories:
    • Remote sensing: satellites, aircraft; passive and active depending on sensor type.
    • Proximal sensing: ground-based vehicles or low-flying platforms; active sensors illuminate the crop.
  • Advantages of proximal over remote: usable anytime, not dependent on sun/cloud, higher spatial resolution.

Electromagnetic Spectrum and Leaves

  • Visible spectrum and chlorophyll interactions: red and blue absorbed, green reflected; NIR reflects due to leaf structure.
  • Internal leaf anatomy influences reflectance in different bands (e.g., palisade cells, spongy mesophyll).
  • Importance: spectral signatures enable assessment of plant condition and vigor.

NDVI and Spectral Signatures

  • NDVI definition: NDVI = rac{NIR - R}{NIR + R} where NIR is near-infrared, R is red reflectance.
  • Typical range: -1 \,\le\, NDVI \le 1; vegetation tends to higher NDVI values.
  • Use: correlate with biomass; example shows strong correlation (R^2 ≈ 0.82) between NDVI and biomass measures.
  • Other spectral trends: different crops and soils have distinct spectral signatures across bands (Blue, Green, Red, NIR).

Data Applications and Decision Making

  • Biomass estimation: calibrate NDVI to available biomass to set stocking rates.
  • Zoning and variable rate applications: apply fertilizers or inputs variably within paddocks.
  • In-season management: variable rate N application, yield prediction, and targeted weed control.
  • Horticulture example: WeedSeeker targets weeds in fallow; precise spraying reduces chemical use and environmental impact.

WeedSeeker and Fallows

  • Active sensors detect weeds; spray nozzle activates only where weeds are present.
  • Benefits: chemical savings, reduced environmental impact.
  • Disadvantages: expensive to set up; best for fallow with intermittent weeds.
  • Related systems: WEED-IT, John Deere See & Spray.

What You Should Know (Key Takeaways)

  • Understand remote sensing vs proximal sensing concepts and uses.
  • Know how plant sensors work and their production applications.
  • Be aware of precision livestock sensing options, purposes, and challenges.
  • Recognize main challenges associated with various livestock technologies (hardware, data, welfare, power, processing).

Quick Reference Equations

  • NDVI: NDVI = rac{NIR - R}{NIR + R}
  • NDVI range: -1 \,\le\, NDVI \le 1

Practical Takeaways

  • RFID and WOW enable automated, high-frequency individual animal data collection.
  • On-animal sensors enable continuous monitoring of location, movement, and health signals.
  • Virtual fencing offers a boundary mechanism without physical fencing, with welfare and public-perception considerations.
  • Plant sensing (remote and proximal) supports spatial management of crops and rangelands via spectral indices like NDVI.
  • Data analytics (biomass estimation, variable-rate inputs, yield prediction) drive efficient, site-specific management.