Real-time monitoring and smart irrigation with a Farming Automation solution.
The solution
Choosing the right methodology and devices was critical to designing our Farming Automation solution, built on our integrated data, AI + IoT platform. We started by choosing the right controllers and sensors which can automatically adjust irrigation logic to suit crop, weather and soil conditions, and then building out from there.
For the solution to be successful, it needed to apply machine-to-machine communication, backed by machine learning applications that would allow for more precise scheduling decisions.
Defining what data to collect is critical
We started by identifying the key metrics necessary in order to achieve the desired farming outcomes. In this solution, these included:
- Weather station data
- Soil moist probe data
- Irrigation system data
- Irrigation schedule
- ERP data
- Weather history and forecast
- Light forecast
- Production schedule
- Nutrition data
- Sales data
Initial goals of the solution
The first goal of the project was to connect the sensors and systems and push the data to the Rayven cloud. As the irrigation was taking place in a area with bad cellular connection, we first had to set up together with one of our communications partners to launch a private LoRa network.
Once in place, we then needed to make sure that we could send all the data via a LoRaWAN gateway, which would then send that data in real-time to the Rayven cloud. This would ensure that all of the data required to be collected would be accurate and recorded in a consistent and reliable manner, thus ensuring data quality, integrity and provide the following features:
- Monitoring and control of the irrigation system via a web-based and mobile system
- Provide device management monitoring for the irrigation system and sensors in the field
- Enable custom business logic to be programmed into the solution so it could meet differing objectives
- Alarm and alert notifications via email or SMS messages when unexpected issues occur
- Diagnose reasons for downtime
- Enable data modelling based on internal and external data sources, such as weather, moisture and light
- ERP and scheduling information
- Forecast of water schedule recommendation
- The ability to test the application, making sure all of the above goals are met, based on the below solution architecture
Before go-live, we tested four critical aspects of the Farming Automation solution:
Security
Our integrated data, AI + IoT platform is built with security as a top priority and our proprietary security architecture ensures that data is secure at all points of the environment.
The solution includes data encryption in transit from device-to-cloud, as well as device authentication; security (Bearer) tokens; SSL, AES and RSA encryption; as well as additional device security checks done via automated polling.
Connectivity
As irrigation is in a remote area with bad cellular connection, we couldn’t use 4G (or 5G) to collect data. Instead, we had to set up a LoRa network with one of our communications partners and use a Industrial IoT LoRaWAN Gateway with a 11.5 KM range to collect data from the sensors and systems in the field, before then sending the data to the cloud with a 4G Cellular LoRaWAN Gateway.
Data Integrity
As we were using data from very different systems, all in remote areas with limited connection, it was critical for us to test the connectivity to minimise the issue of data package lost. We also ran testing of data back-fill, to make sure that if there is a communication issue, we can fill the blanks in data accurately so that there is never a data loss issue that can affect data integrity.
Industrial data science
The objective of exploratory data analysis was to observe trends in the soil, weather and irrigation data, which included:
- Real-time forecasting of evaporation rate based on weather data
- Irrigation forecasting based on evaporation rate
- Forecast water usage based on weather data
- Automatically update irrigation schedule based on forecasting
- Irrigation system predictive maintenance
- Water tank level forecasting
What’s next?
We are continuing to collect data and provide ongoing insights to our customer, as well as continuing to work on improving even the forecasting results even further by utilising machine learning as the data sets grow over time.
