Prototype of an Intel FPGA-based automatic irrigation system for soft fruit farms in Perthshire, Angus and Fife, Scotland, UK

Demo Video

Project Proposal

1. High-level project introduction and performance expectation

The purpose of the design is to reduce water waste, efficiently irrigating crops at soft fruit farms by automating the process and protecting crops from the growth of fungus by using less water. This is achieved by only using water as required, instead of by timed irrigation, which is not responsive to climate conditions. Using the optimum amount of water is economically beneficial as it speeds up the ripening and consistency of the crop. It also reduces the chance of harm and pollution to the local environment.

Initially, the target user-base is soft fruit farmers in Perthshire, Angus and Fife, Scotland, UK. These farms grow soft fruits such as cherries, blueberries, aronia berries, strawberries, raspberries and redcurrants. Farms in different regions of the world which are growing other crops would also benefit from the Intel FPGA-based (Field Programmable Gate Array) irrigation system described below and, although not in scope of this project, a future iteration of the system could be readily adapted to manage the delivery of drinking water to livestock.

The application scope consists of an Intel FPGA-based irrigation system, a Terasic DE10-Nano SoC (System on Chip) Cloud Connectivity kit. Instead of timed or manual starting of the operation of a water pump, the Intel FPGA-based system will irrigate only when required by measuring the level of moisture in the soil (pH acidity levels and temperature measurements may also be collected). An ADI (Analog Devices Inc.) shield with a capacitive soil sensor is used, producing an analog voltage signal, the value of which depends upon the amount of soil moisture. The signal is fed from the sensor into an Analog-to-Digital Convertor (ADC). When an appropriate threshold is reached, the ADC provides a digital output signal suitable for processing by the Intel FPGA, which sends an output signal to a relay to switch on a water pump if the soil is dry and then switches off the water pump when the crops have the necessary amount of water. This process cycle is repeated for the duration of the growth of the crops.

The Intel FPGA Cyclone is ideally placed for this solution with its attributes of low power, low cost, and flexibility. For enabling the steps of coding, simulation, synthesis and compilation, the powerful and versatile software-programming design application Intel Quartus Prime, which includes an implementation of the Hardware Description Languages (HDL) VHDL and Verilog, can be used to configure the Intel FPGA.

Development time can be reduced with Intel Quartus Prime, which can also be used for correcting errors, carrying out timing analysis and editing logic circuits and diagrams using the built-in visual tools. Modelsim and its Graphical User Interface (GUI) can be used in conjunction with Intel Quartus Prime to enable simulation, verification and debugging for the (HDL) language used.

If time allows, an ADI wi-fi module may be used with the Intel FPGA-based sensor system, to connect from the network edge IoT (Internet of Things), i.e. the farm, to Microsoft Azure cloud services. SaaS (Software as a Service) capabilities in Azure IoT Central mean building a cloud application from scratch may be avoided by using the app templates provided by Azure IoT Central, where deep technical knowledge is not required. The potential to send and receive data to and from Azure, such as settings and telemetry, allows insight, monitoring and the ability to develop relevant farming apps and solutions.

The Terasic development kit based on an Intel Cyclone V SoC (System on Chip) FPGA, comprises of a Hard Processor System (HPS) portion and a FPGA portion. Embedded Linux running on the HPS will be used to connect to Microsoft Azure and connectivity through the ADI shield to sensors is via the FPGA fabric. Also, a Terasic RFS sensor expansion board connects to the FPGA via GPIO connections.

2. Block Diagram

3. Expected sustainability results, projected resource savings

The capacitive soil sensor outputs a varying voltage dependent on the level of moisture in the soil, varying from a low range when dry, to medium when moist, to a higher range when saturated. Initially, it may be required to experiment and collect data in the field to determine the low, medium and high points that the soil sensor outputs. The analog voltage from the sensor is amplified, then fed into an Analog-to-Digital Convertor (ADC) and converted into a digital signal which is outputted from the ADC to the Intel FPGA. The ADC is to be calibrated and lower and upper thresholds used in deciding when to activate the switch relay after the signal is processed by the Intel FPGA. The relay either switches on the water pump, or switches off the pump. The water pump will be inactive below the upper limit of the dry threshold and activated above the lower threshold of either the medium or high ranges. This gives flexibility to the farmer to adjust the system to deliver the optimum amount of water as the crops may require different amounts of water depending on season and stage of the crop’s growth, thereby reducing waste and minimising any impact to the local environment. The pump only operates when required, saving resources - energy (electricity), labour and water.

4. Design Introduction

The purpose of the design is to reduce water waste, efficiently irrigating crops at soft fruit farms by automating the process and protecting crops from the growth of fungus by using less water. This is achieved by only using water as required, instead of by timed irrigation which is not responsive to climate conditions. Using the optimum amount of water is economically beneficial as it speeds up the ripening and consistency of the crop. It also reduces the chance of harm and pollution to the local environment.

Initially, the targeted users are soft fruit farms in Perthshire, Angus and Fife, Scotland, UK. These farms grow soft fruits such as cherries, blueberries, aronia berries, strawberries, raspberries and redcurrants. Farms in other regions growing different crops would also benefit from the Intel FPGA-based irrigation system described below.

The application scope consists of an Intel FPGA-based irrigation system. Compared to ASICs, an FPGA-based irrigation system reduces risk for field trials by reducing development time to produce a working system, whilst allowing for the ability to quickly make changes before any volume production.

The Terasic DE10-Nano FPGA is ideally placed for this solution with its attributes of low power, low cost, and flexibility. For enabling the steps of coding, simulation, synthesis and compilation, the powerful and versatile software-programming design application Intel Quartus Prime, which includes an implementation of the Hardware Description Languages (HDL) HVDL and Verilog, can be used to configure the Intel FPGA.

Development time can be reduced with Intel Quartus Prime which can also be used for correcting errors, carrying out timing analysis and editing logic circuits and diagrams using the built-in visual tools. The multi-language environment Modelsim with its Graphical User Interface (GUI) can be used in conjunction with Intel Quartus Prime to enable simulation, verification and debugging for the HDL language used.

Bench Setup of DE10-Nano FPGA, RFS expansion board, shield and sensors:

5. Functional description and implementation

The functionality of the design of the Intel FPGA-based irrigation system means that, instead of timed or manual starting of the operation of a water pump, the Intel FPGA system will irrigate only when required by measuring the level of moisture in the soil. A capacitive soil sensor is used producing an analog voltage signal the value of which depends upon the amount of soil moisture. The signal from the sensor is amplified and fed into an Analog-to-Digital Convertor (ADC). When an appropriate threshold is reached the ADC provides a digital output signal suitable for processing by an Intel FPGA which then sends an output signal to a relay to switch on a water pump if the soil is dry and then switches off the water pump when the crops have the necessary amount of water. This process cycle is repeated for the duration of the growth of the crops.

6. Performance metrics, performance to expectation

More than one field may be controlled by the Intel FPGA irrigation system and these may have different watering requirements. To maximize crop yields and minimize waste the Intel FPGA may be programmed to allow for different outcomes for different fields.

The size of the soft fruit farm also determines how the Intel FPGA irrigation system may be scaled in size. Several presumptions may be made in the next stage of the contest with regard  to the scope and size of the system.

The Intel FPGA outputs in the program require to be tested to correct any significant errors, fluctuations or noise that may affect the output signals prior to downloading the design to the Intel FPGA.

The benefits of using Intel FPGA devices in the design include low cost, flexibility, availability and potential use of the powerful Intel Quartus Prime design software, programmability and reusability of the FPGA devices. Also, if there are design changes required, it is quick to make the changes. The adaptability of the Intel FPGA makes it ideal to be the central component of this automatic irrigation system.

Performance metrics also include comparing crop harvests before and after implementation of the FPGA system, ie crop yield per unit area of land. Also, measured water usage would be expected to optimized.

The expected performance after implementation is expected to be stable on a long-term basis. Alerts may be setup on Microsoft Azure to report on any failing sensors or boards so that any maintenance required may be carried out in a timely manner.

7. Sustainability results, resource savings achieved

The implementation of the system boosts performance and increases efficiency by automating the irrigation process, leading to significant resource savings in time and labour costs, reduced power costs for water pump operation, decreased water usage and increased crop yields.

Possibly the number of Input/Output (I/O) interfaces of the system can be expanded with multiple, stacked ADI shields (ADI EVAL-CN0398-ARDZ shield), although this has not been tested by the project. This potential flexibility means that more sensors may be added.

Also, depending on requirements, the system can be adapted to change and in conjunction with adjustments to the underlying farm LAN/Wi-Fi network, a larger area of farmland may be serviced by automated irrigation using additional Terasic DE10-Nano kits.

The FPGA may be powered by an appropriately sized small solar panel and battery backup, reducing reliance on mains power, again reducing installation and running costs as mains power can be difficult to install in outlying farm fields.

More predicted sustainability results and resource savings are summarised below:

Water

- The implementation of our irrigation system saves the farmer valuable resources including water, time, and manual labour.

- As the crops are watered close to optimum level, it is likely that there would be less time needed sorting out the spoilt crops after harvest.

- Over a period of many years, this project has the potential to save a huge capacity of the farm’s water.

- Saving water resources boosts the farm’s overall sustainability.

- Over time and with more sensors added, the farm can be fully optimised in its water usage.

- Multiple sensors can help scientists and farmers understand the hydrological and soil dynamics of their cropland, which can lead to smart water usage decisions at the watershed scale.

- Overwatering is a constant strain on the soil system.

Soil Health and Quality

- Optimal water usage prevents soil erosion on the farm. Therefore, ensures livelihood and sustainable farming practices over time because the soil health will be kept in good condition in the future.

- Soil erosion is a serious long-term threat to farmers.

- Soil erosion is a serious issue impacting farming sustainability due to the change in soil composition and the transportation of important nutrients out of the field. This leads to a decline in crop quality over time.

- Overconsumption of water releases the soil’s organic carbon which contributes to climate change, the use of our system will help prevent/decrease this and means the farming can be more sustainable and last longer.

- With regard to soft fruit farms (also applies to other types of farms and crops), over-watering and the decline in soil health, leads to the release of greenhouse gases (GHGs) from the soil and decreases the net primary productivity of the soft fruit plants. This decrease in net primary productivity negatively impacts fruit yield. Our system would help prevent this, by offering a more sustainable route to the maintenance of soil quality and the carbon storage potential, which helps climate change mitigation.

- Good soil quality means a decrease in flooding potential, allowing the farm to sustainably continue production and not waste crops due to flood events.

- Less flooding means less pollutants and contaminants in the fields.

- Prevention of flooding by healthy soil and the understanding of hydrological parameters through our system saves money, reduces environmental and economic costs due to the promotion of good water and soil health. Our system encourages optimum biogeochemical cycling within the farm’s soil, and ensures sustainability of the farm.

8. Conclusion

To implement the Intel FPGA-based solution in this proposal by automating and optimising the irrigation process leads to a reduction in water wastage and likely increases in crop yields. This is a real benefit to traditional farming and to innovations such as vertical farming, particularly in arid areas of the world where water resources are already scarce.

The FPGA Intellectual Property developed for this project may be available as Open Source material for Sustainable Agriculture, potentially enabling low-cost deployments at scale by service providers or Non-Governmental Organisations. This is particularly relevant to rural and low-income regions of the world currently unserved by existing internet infrastructure - near future rollouts of new technologies, such as SpaceX's Starlink or Meta's aerial fibre project, will allow these regions access to the internet and thereby to Microsoft Azure cloud services making this Intel FPGA-based solution a powerful and valuable tool, thus playing a vital role in mitigating the effects of the ongoing climate emergency.