The system gathers real-time data from moisture and temperature sensors and 2 hours forecast weather data detail by using Weatherapi.com. First, I build this with intent for normal soil type and common plants type. The condition is if the soil moisture is less than 30, the soil temperature is between 20 and 30°C, and the forecasted precipitation raining amount is less than or equal to 2 millimeters, then a watering action is performed. This approach enables farmers to maximize the use of limited resources, as the system automatically adjusts watering based on specific conditions. Additionally, there is overwrite forcing setting to water and stop watering for various purposes such as maintenance, pesticide, or fertilizer. Next, a water flow sensor is deployed to monitor and detect any potential water leakage within the system.
In implementation state, I will use Azure IoT Central which require subscription. To continue, I will sign in to Azure Hub then, move to IoT Central application and create new application.
After filling in the required field, choose custom application for device template and create the application.
Then click “IoT Central Application URL” to move to the application page.
On application page, go to device templates section and create new template by selecting IoT device panel.
Next, added requirements file that relevant to smart irrigation systems. Here, I will add different types of state and command for different purposes.
An overview for data visualization has been developed to provide a better understanding of the information generated by the smart irrigation system.
After that, I will create a device with irrigation template and connected with python program for data stimulation and receiving command, etc.
All specified conditions, including soil moisture, air temperature, and precipitation forecasts are evaluated to make automated decisions based on real-time and forecasted weather data.
Users can perform action to overwrite system automation in command section.
Upon activation of a command, the corresponding response can be retrieved from the terminal. Additionally, the impact of the executed command becomes visible in the overview, allowing users to observe and confirm the changes made within the smart irrigation system.
Furthermore, users have the capability to schedule jobs in the designated Job section to schedule forced action. Users can establish rules to receive email alerts in the event of system abnormalities, providing an effective means of proactive monitoring and timely response to potential issues within the smart irrigation system.
To assess the effectiveness of my prototype, I employed a multifaceted evaluation approach using both simulated scenarios. Controlled environment testing: I set up the prototype with simulated weather conditions and soil moisture levels. This allowed me to test specific functionalities under predictable conditions. Pilot deployment on select location: I stimulate the system on different geographic location, including their historic weather data and common type of plant they farm. This provided valuable data on the system's performance in various field conditions and interactions with historical irrigation infrastructure. The prototype effectively addressed the key problems identified earlier, specifically inefficient water use, lack of real-time monitoring, and non-adaptability to diverse conditions. Evidence from the evaluation process highlighted the system's ability to optimize water distribution based on real-time data, adapt irrigation practices to varying soil types, and cater to the specific moisture and temperature needs of different plant types. User satisfaction in the control features, and the system's adaptability to unexpected weather changes showcased its resilience.