Precision agriculture is big business – and it’s about to get even bigger. According to GPS World, the global market forfrom $4 billion in 2019 to $12 billion in 2025.
The principle behind precision agriculture is simple. The more that farmers know about the land on which they plant their crops, the more efficient they can be when it comes to planting, cultivating and harvesting them – and that can deliver a major boost to farm profitability.
By applying the right treatment at the right time to each plant, for example, farmers can make huge savings on inputs like water, chemicals, fertiliser, labour and fuel. In 2016, a US Department of Agriculture survey found that corn farmerslike yield mapping and variable rate technology (VRT).
Four types of precision farming technology
While the premise of precision agriculture is simple, the technology that enables it is varied, complex, and evolving all the time. Broadly speaking, there are four types of technology available today:
Guidance and steering systems: Semi-autonomous and autonomous steering systems for tractors and other equipment, allowing the vehicle to navigate along a pre-set path within the field. This means the driver can focus on the cultivation work at hand, rather than on steering the vehicle. Semi-autonomous vehicles are becoming commonplace in agriculture, and a new generation of completely autonomous, driverless vehicles is fast emerging (although the challenge of driverless navigation between field and farm – especially on public roads – is yet to be addressed).
Land preparation: There are many technologies available for mapping farmland and sampling the soil to understand its chemical make-up and variations. With detailed knowledge of soil characteristics across the farm, farmers can more efficiently plan how best to prepare the soil, what to plant where, and how much water and other inputs each area of crop will need.
Yield monitoring and mapping: Monitoring the quality of crop yields across the farm can also provide invaluable data for future planning. Monitoring technologies generally consist of location-aware sensor-based systems that record yield quantity from each location as it is harvested, with the data then mapped into a farm analytics and management system.
Variable rate technology: Technology that automatically alters the rate at which seeds are planted or watered, and chemicals and fertiliser are applied, based on the specific conditions at the location of each plant. Variable rate technology is either pre-programmed using mapping software and positioning technology, or applied in real time using sensors to guide the application.
The future of farming is AI-driven and autonomous
These technologies are already evolving to embrace the use of AI and autonomy for further efficiencies and cost reductions. For example, soil and yield mapping data can feed into predictive systems that combine it with weather forecast data and drone footage of the crop to make real-time recommendations for action to ensure the best possible yield.At the same time, driverless tractors, farming robots and agricultural drones are starting to remove humans from the loop when it comes to monitoring, preparing, planting, irrigating, cultivating and harvesting the crop. These technologies rely on a multitude of sensors combined with accurate positioning systems to automate work previously conducted by humans.
Testing positioning performance is critical
For precision agriculture technologies to deliver on their promise of more efficient and profitable farming, they must be fit for purpose. For the companies developing and integrating these technologies, that means rigorous testing to ensure their products perform to expectation on every farm, in every location around the world.
One of the most critical functions of a precision agriculture system – whether a tractor, drone, or robot – is positioning. The system must always know exactly where it is, so it can provide an accurate soil map or apply the right amount of water or fertiliser to the right plant.
In precision farming, position accuracy is required down to the centimetre level, creating a requirement for high-performance positioning capabilities. This kind of accuracy can’t be obtained with GNSS alone, so precision farming equipment fuses GNSS data with correction data like Real Time Kinematic (RTK) and Precise Point Positioning (PPP), as well as data from other sensors on the vehicle, like inertial measurement units (IMUs), lidar and camera vision.
Simulation is the best way to recreate a wide range of real-world conditions
Assessing the performance of these multi-sensor positioning systems means subjecting them to a wide range of conditions to ensure they perform as expected in each case.
That’s most efficiently done with simulation. Using simulators, developers can create many different conditions that the system may encounter in day-to-day use on the farm, as well as unusual ‘corner cases’ that may happen rarely, but can have serious consequences if they do.
For positioning systems, such conditions might include:
Atmospheric effects: Atmospheric effects like space weather and ionospheric scintillation can affect the accuracy of the GNSS signal, with more pronounced effects in certain regions of the globe (for example, equatorial countries like Brazil). Certifying that the receiver can cope with atmospheric effects specific to different regions can be a good competitive differentiator.
Multipath effects: Satellite signals may reflect, refract or diffract off objects in the environment (including the ground), fragmenting the signal and resulting in inaccurate readings if not compensated for in the receiver design. While multipath is not as much of a problem on open farmland as it is in urban locations, it can still be caused by rotating equipment mounted on the vehicle. Buyers of precision farming systems will want to be sure their performance won’t be compromised by multipath signals.
Radio frequency interference (RFI): There are many ways the satellite signal can be degraded through interference, sometimes to the point where it becomes unusable. Man-made jammers can affect signal reception, as can activity in adjacent frequency bands (for example, radio and TV transmitters), noise from poorly-isolated elements within the receiver circuitry, and noise from other RF equipment placed on the vehicle. Precision agriculture receivers should be as resistant as possible to RFI, which means assessing their performance in many different interference scenarios.
Signal spoofing: Criminals and other malicious actors have started to use RF equipment to ‘spoof’ GNSS receivers by broadcasting a false GNSS signal from a transmitter on the ground – drowning out the real signal and forcing the vehicle to lock on to the fake one. This can be used to take control of an unsupervised driverless vehicle, forcing it to navigate off course – and into the hands of thieves. While still rare, a growing number of real-world incidents show that this is an emerging threat to autonomous vehicles, and one that manufacturers and integrators would be wise to test for.
Error correction: Many precision agriculture systems rely on error correction services like RTK and PPP to increase the accuracy of the GNSS signal to centimetre level. Receiver developers and integrators must verify that error correction algorithms are correctly applied within the receiver, which can be done efficiently through simulation.
Sensor fusion: Where multiple sensors and inputs are combined to generate a position, manufacturers will want to be sure that the algorithm correctly combines the different inputs to compute a position that is accurate. Sensor data simulation software like Spirent’s SimSENSOR, combined with a GNSS signal simulator, can provide a professional testbed to characterise the performance of sensor fusion algorithms.
Multi-constellation, multi-frequency receivers and future signals: Many precision agriculture systems use receivers that process signals from more than one global navigation satellite constellation on more than one frequency – both as protection against signal jamming and individual constellation failures, and to deliver a more accurate position (as more satellites can be used). Simulation of multiple satellite constellations is an efficient way to assess the impact on the receiver if one constellation fails (asin 2019). Planned future signals can also be simulated, to test receiver performance in advance of the satellites becoming operational.
How Spirent can help
Spirent’s professional services teams work closely with receiver developers and integrators to help them design testbeds and assess the performance of positioning systems for use in precision agriculture.
In addition, we offer a wide choice of simulation and other lab test equipment and software, including the market-leadingand PosApp control software, and RF interference generators, and specialist simulation software for realistic multipath modelling ( ), sensor data (SimSENSOR), error correction data (RTCM), and spoofing ( ). Through our powerful API, we also integrate with many real-time vehicle dynamics simulators for added motion realism.
Spirent's, meanwhile, allows you to record the full spectrum of RF activity in any region of the world, and play it back in the lab – allowing for repeatable testing with real-world signals. Alongside the RF environment, the GSS6450 will also simultaneously record data from inside the vehicle including IMU data, CAN bus data, and camera vision feeds – as well as data from error correction services.
With Spirent as your trusted test partner for positioning, navigation and timing, you can develop innovative, reliable and desirable solutions that allow customers to move forward with confidence into the new and profitable world of precision agriculture. To learn more, visit Spirent'sor .