Designing open-source smart farming systems - Embedded.com

Designing open-source smart farming systems

In the last 20 years, computing has dramatically changed the world, and the agriculture industry has felt the impact. Advanced equipment can now lead to more efficient production; modified seed can make pest management easier; and automation has allowed many chores to change, even if they don’t disappear. However, farmers have increasingly found the technology they rely on to be a double-edged sword.

At the same time, farmers struggle against a host of mounting business challenges. They also find themselves unable to service or repair their machinery in the same ways they used to because of software digital rights management and other equipment design decisions that favor the service revenue of equipment manufacturers over farmers. At the most extreme reaches of this tension between suppliers and farmers, farmers have even become beholden to exclusive producers of seeds for some crops, as cross-pollination with patented, genetically modified plants in adjoining fields draw litigation from seed producers.

These struggles leave farmers around the world reaching deep to remain competitive. In this competition, they find themselves being pushed by suppliers toward seeds, equipment, chemicals, and other inputs into their business over which they have incomplete control. Farmers are losing the choice to repair their equipment, save seed, or take other time-tested, cost-saving measures. They are wary of the unmentioned costs of technology in their work.

With this challenge comes an opportunity to build agricultural tools that give power back to the farmer. If farmers can build or retain the right to edit and use more of their equipment, the path toward “smarter” farms becomes much more appealing for farmers. Agriculture in general is seeing massive opportunities for leading-edge technology such as the internet of things and artificial intelligence. Open source and other serviceable means of implementing these tools is critical for farmers to maintain their competitive edge.

Designing open-source, low-cost smart farm systems

A recent research project conducted in Bangladesh has built on previous work by creating an open-source model for smart farming. This model makes heavy use of Arduino components, lays out a comprehensive environmental monitoring system, defines approaches for data analysis, and creates paths for automated farm input management. While the primary focus of the test system built was field production, it is already anticipated that the general architecture could be applied to greenhouses, livestock production, and more agricultural applications in the future.

Basing all monitoring equipment on the Arduino Mega 2560, this project centered on a mesh of “monitoring nodes” set throughout an agricultural field, with a single “central node” aggregating all data, triggering automated events, and passing data to the cloud.

Building a sensor mesh

Each “monitoring node” adds an array of sensors to the Arduino Mega board. The researchers involved chose devices that were accessible for wide demographics based on cost, ease of use, and flexibility.

To outfit a single monitoring node, the Arduino board was extended with the following:

  • nRF24L01+PA/LNA wireless transceiver module for communication back to the central node — Using an embedded wireless protocol and 2.4-GHz spectrum, this device can transmit as far as 1,100 m; the 2.4-GHz spectrum allows for 125 unique channels, extending the number of connections possible for a single farm and central node.
  • DHT11 temperature and humidity sensor for surface temperature and humidity measurement.
  • PIR motion sensor, which depends on infrared imaging to detect motion for alerting farmers to pests or trespassers.
  • MQ-135 gas sensor, which detects several organic compounds in the air, including smoke.
  • BMP180 barometric pressure sensor to detect changes in barometric pressure, indicating the potential for storms or other incoming weather patterns.
  • Soil moisture sensor, which measures electric resistance within the soil to determine the water saturation level of the ground.
  • Rainfall detection sensor, which checks for falling rain by measuring resistance across a sensing plate.
  • pH meter, which determines soil acidity and potential need for soil additions.

With a wide selection of attachments in hand, the research project created several sensing units that could be deployed across a large field. Given the wireless communication hardware selected, a single central node could serve dispersed monitoring nodes in up to a 2,200-m diameter. To prepare each monitoring node, each Arduino was wired with its sensor array, programmed to store and share all sensor data and addressed for the sake of communications.

Once monitoring nodes were created, the team turned its attention to the central node, tasked with further analysis and communication of field conditions.

Equipping communication and farm management

Overseeing the collection of monitoring nodes is a “central node” designed to aggregate, report, and act on the data collected from around a field. Built on the same Arduino Mega 2560 platform, it relays what is happening across the farm. In place of the sensor equipment packaged with each monitoring node, the central node has a Wi-Fi card and cellular radio with a SIM card for “always on” transfer of data to the cloud. The research team further conceptualized adding servos, motors, and similar hardware if and when it would be useful for management automated irrigation and other complementary systems across the test farm.

With this design, a single central computer and GSM communications device can relay information collected over a large swath of land. Other advantages are numerous — given its centralized functions, the central node is programmed to:

  • Alert farmers via SMS of changing conditions and needed interventions
  • Manage accessible inputs, such as turning on irrigation to areas needing water
  • Push collected data to the cloud for additional processing, analysis, and action

Tying data collection to total smart farm management

To power the network of nodes, the team paired a photovoltaic solar panel with a battery, allowing for consistent power availability in the field. The basic nature of the components used ensures low power draw from each node, reducing the overall supply required to operate the system.

With a steady stream of openly accessible data flowing from the field, the options for automating more of a farm’s management becomes flexible. As an example, the research project chose Google Sheets for aggregation and analysis of sensor data. While Google’s software is well suited to this purpose and Google suite, in particular, is free, many data storage and analysis tools could serve the same function.

On the topic of purpose-built agriculture drones, the team similarly was able to create a basic quadcopter with off-the-shelf components. Specific alterations to a typical drone will eventually include the ability to carry and precisely apply payloads of pesticides and fertilizer. Further improvements in time may include advanced approaches in pest detection and elimination, additional crop data collection, or could even take advantage of machine-learning advances in grading produce.

As each of these pieces fall into place, a high-performance, open concept for managing the entire farm takes shape. While additional considerations such as security are not explicitly addressed at length in this concept, the open nature of the design allows for any prospective builder to consider this and other concerns about managing their own data. With control over their purpose-built “internet of things,” farmers may be able to enjoy the improvements to insight and efficiency IoT promises without losing serviceability.

What comes next

While this research project was intended to add to a body of work on affordable access into smart farming, the value extends further. The concept of using popular and widely available components such as Arduinos does not only reduce the cost of entry into smart farming; it also means farmers have more control over their business.

Particularly for subsistence farmers and medium-sized operations the world over, this democratization is happening at a critical time as they navigate difficult business decisions.

>> An earlier version of this article was originally published on our sister site, EEWeb.

 

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.