Automating Agritech with a Sensor Inventory Management System

Requirements gathering

The development of the sensor management system for farms began with a deep dive into the business challenge — automating the tracking and management of irrigation sensors. Our software engineers worked closely with the customer’s operations team to define their needs and understand how technology could enhance their existing process of sensor inventory management. The primary objective was to streamline inventory tracking, improve sensor lifecycle management, and enable close to real-time data updates.

System architecture and design

The next step was to establish the system infrastructure, including setting up the backend architecture to support integration with external platforms for sensor deployment, retrieval, and monitoring. When designing the system, we used REST APIs for smoother data exchange between the system and other platforms. This also added to flexibility and scalability for future updates.

We structured the core functionalities around the following key components:

• Sensor analytics: This component processes sensor data streams to analyze them and generate useful insights and data products.
• Sensor workflow management: This component automates the deployment and retrieval process to make sure that inventory management sensors in agriculture are properly tracked and actions are triggered when specific tasks are completed. It also integrates with several external systems to manage sensor workflows effectively.
• Sensor reporting: This feature offers an overview of sensor performance and status in a user-friendly format. It supports operational reporting, allowing users to learn the sensor status (installed vs. warehouse), its location, and potential performance issues.
• Service infrastructure: This part of the system provides the foundation for all core functionalities. On the one hand, it helps with the smooth operation of the sensor service. On the other hand, it contributes to easy integration with external platforms and systems.

The diagram below illustrates the sensor data journey, highlighting the key steps in the data flow — from data collection to sensor analysis.

Mobile integration

A critical part of the system was integration with a mobile application for field agents. This allowed agents to log in and update sensor information in close to real-time while on-site, keeping the data current and reducing manual updates. This way, field agents could manage their tasks more easily while the operations and commercial teams were able to track sensor statuses.

Bulk sensor management

To speed up the workflow and add more convenience, our team implemented bulk sensor management. This feature allows users to add sensors in batches using CSV uploads. It also works for batch editing to save the teams’ time spent cataloging and updating sensor information.

QA/QC checks

Maintaining data accuracy and sensor reliability was essential, so we integrated automated quality control (QA/QC) checks at key stages of the workflow. These checks helped detect and address potential issues early in the workflow to minimize the risk of faulty data entering the system. Specifically, we checked whether:

• The sensors for inventory management were correctly deployed
• The right sensor was used for a specific location
• No data was lost or corrupted during the transfer or collection process
• The sensor did not produce any data at all
• There were any issues with uploading sensor data after deployment

System integration and final implementation

With the core components in place, Intelliarts worked on deploying the system and integrating it with external platforms for sensor deployment and retrieval. Our engineers optimized the software for scalability so it could handle larger volumes of sensor data and accommodate new sensor models when this need arises.

The software solution for farm inventory management with sensors delivered by Intelliarts brought significant improvements to operational processes. By addressing key challenges, such as manual data uploads and inventory tracking, we helped improve the company’s ability to manage irrigation practices and scale their impact.

Here are the key measurable results we achieved:

• Efficiency in sensor management: The system helped streamline agricultural warehouse management with sensors and reduce time spent on manual tasks. As a result, the customer is on track to scale operations by more than 260% in the near future.
• Sensor data accuracy: Better data management, field coverage, and a more robust QA/QC increased overall data accuracy and reliability. For instance, the company experienced a 10x reduction in water use per acre as compared to the prior year. Another notable improvement was the detection of fewer irrigation events on specific fields after the system was implemented. In contrast, unreliable sensor data in the previous year forced our partner to use default assumptions about irrigation practices.
• Automation: The customer eliminated manual data uploads by building the automated system. This shift helped get rid of complex SQL databases, inconvenient Google Sheets, and ad hoc dashboards. In the long run, it also improved the operations team’s productivity by allowing them to focus on higher-priority tasks.
• Scalability: With the new system in place, the customer got an opportunity to expand their operations and serve extra fields without compromising data quality or process predictability.