Banana Plantation Monitoring & Automation
  • Solution Deployment Size:

      30+ sensors
  • Solution Type:

      Subscription Based
  • Key Business Requirements:

      Monitor & analyze mission-critical banana tree related data across the entire plantation
      Creation of automatic warning system to prevent human error
      Reduce need for hiring costly experts
  • Other Usage Categories:

      Palm Plantation
      Pineapple Plantation
      Plantations
  • KPIs

      % Increased Productivity – 5-8%
      % Reduction in Cost of Labor – 10-30%
      % Reduction in Loss of Produce – 12-28%
      Increased Quality of Product
  • Data Monitored:

      O2 (Soil), Humidity (Soil), Humidity (Air), Temperature (Air), Temperature (Soil), Solar Radiation (Air), NH3 (soil), Fruit Diameter, Tree Trunk Diameter
  • Solution Tags:

      Preventing soil flooding, Disease control, Growing Fruit Efficiently
  • Regulations Addressed by Solution:

Business Case / Problem:
  • Banana trees are extremely vulnerable to disease and plagues in the winter season
  • Excessive use of pesticides and overuse of other traditional methods can cause severe damage to the quality of product
  • Determining the root causes and prevention of the above issues using manual/traditional methods requires a very large amount of labor
Solution Overview:
  • Soil Humidity, temperature and oxygen content is tracked across the entire plantation in real time through deployment of low-cost, low-energy devices to catch environmental problems before they cause disease and plague
  • Luminosity across the plantation is gathered using luminosity sensors to identify specific areas of the plantation at higher risk
  • Data is used to create automatic warnings and alarms for operators using Thing+’s rule system via email & SMS
  • Warnings & alarms send the exact location of the issue to operators where they can check real time and historical data of that area of the plantation
System Diagram & Hardware Used:

(Package) Libelium Waspmote Plug & Sense! Smart Agriculture

(Package) Libelium Waspmote Plug & Sense! Smart Environment

  • Other Possible Hardware Vendors:

      AdvanWISE
      Advantech B+B Smartworx
      Maestro Wireless
Installation Process Overview:

1. Hardware selection and review of the plantation’s unique characteristics.

2. Configuration of sensor nodes and gateway.

3. Placement of sensor devices out in the field – sensors are placed in the soil / and connected directly to trees.

4. Thing+ site and dashboard is set up. User rights and roles are created.

5. At this point the Thing+ platform based monitoring dashboard is set up for operators and stakeholders to view and work with sensor data.

6. Automatic warning / automation rules are set up in the system for detecting potential flooding, excessive humidity, etc….

7. Final checks of the dashboard, rules and data analysis choices are done. Operators at the site are educated on how to monitor and use the system. In this case, operators were also educated on how to add new rules if required for automation of the SMS/email/etc.. warnings.

Automation & Rules Guide :
Example Rule #1: Send SMS based on soil humidity & temperature

Trigger: Temperature is below a dangerous level

Condition: Soil humidity is high

Action: Send SMS message to operator including sensor & location

Example Rule #2: Send SMS & email to operator based on low sustained luminosity levels and high humidity

Trigger: Luminosity is below average

Condition: Soil humidity is high

Action: Send warning SMS and email with data

Impact on Operations:
  1. Operators should monitor data across the plantation to determine locations that are at risk of developing disease and plague. Critical data includes humidity, temperature, solar radiation, and soil NH3 content – daily monitoring is crucial, especially during extreme weather
  2. Solar radiation levels across the plantation should be monitored to identify areas that are not getting sufficient amounts of light
Recommended Data Analysis Options:

Example Data Analysis #1

Humidity Analysis

Data: Humidity across the entire plantation

Frequency of Analysis: 7-30 days

Goal: To be used and analyzed to determine overwatering, underwater, and other potential water usage and flooding problems

Usage:

  • Humidity levels will vary across the plantation and will indicate what areas are receiving optimal levels of water
  • When/if flooding occurs, humidity levels can be used to determine what parts of the plantation are not draining properly

Example Data Analysis #2

NH3 Analysis

Data: NH3 concentrations across the entire plantation

Frequency of Analysis: 7-30 days

Goal: To optimize the amount of fertilizer used.

Usage:

  • NH3 concentration should be used to determine areas of the plantation that are being under or over fed
  • Extremely high NH3 concentrations can cause health issues for plants and result in dead trees
  • NH3 concentration analysis should also be used to determine areas where NH3 concentrates due to excessive humidity/rain

Example Data Analysis #1

Trunk and Fruit Width Analysis

Data: Widths of tree trunks and fruits across the plantation

Frequency of Analysis: 5-30 days (based on season)

Goal: To understand how quickly different areas of the plantation are growing and to inform harvesting projections before harvest.

Usage:

  • Fruit and trunk width data should be used to understand the approximate size of the harvest prior to sale – can be used to set supplier and trader expectations as well as to understand in detail how well trees and fruit are growing in real terms
  • Fruit and trunk widths can also be used to identify problems with the trees and fruit before it occurs, catching many forms of parasite and disease before serious issues can occur
Solution Assumptions:
  • A 2g, 3g, LoRaWAN, LoRa, or other WAN connectivity should be available near the gateway’s location.
Support & Maintenance Requirements:
  • Sensors may need to be recalibrated once per year
  • Sensor node devices should be checked at the same time for signs of aging/problems with solar panels
  • Farms may need assistance with analyzing their data. In this case, support should include analysis of data and providing actionable information to farmers.

Providers/Installers of this Solution:

Solution Extensions:

  • A drone-based sensor platform could be added in order to supplement ground-based sensors with aerial photography and topography to optimize growth, fertilizing, picking and watering further
  • Automated banana picking machines could be integrated into the system’s fruit size data to determine optimal times to pick fruit