Arable farming: Top 5 technologies
Last Week, we introduced the concept of ‘precision agriculture’. This term describes a range of practices that use technology to provide crops with precisely (no more and no less than is necessary) the amount of nutrients and treatments needed for a high yield .
Automation avoids over-application of agrochemicals which saves money and reduces spill over to the wider environment (eg nutrient leaching) and predictive technologies can help customise crop varieties for improved performance . Here are some further examples of the practices that improve efficiency and sustainability in today’s farms.
1. Remote soil and plant sensing in site-specific applications
Sensing equipment attached to vehicles can measure a wide range of soil parameters such as moisture content, composition, density, and can pinpoint the precise locations of individual plants. The results of these measurements help determine how much water or fertilizer are needed and where they should be applied . Seeders, spreaders and sprayers (drones are beginning to be used for this) equipped with variable-rate technology can then supply only what is needed for the growing crop  . Another advantage of soil sensing is that it can monitor conditions in the field without stressing the soil (for example, by compacting or mixing).
Sensing is not only useful in crop husbandry. A remote technology called near infrared reflectance (NIR) is a non-invasive technique that measures the reflection of different wave light lengths to monitor the composition of grain as it ripens (moisture, starch, protein, and oil content) . This data helps determine when to harvest.
2. Machine synchronisation
Combine harvesters already coordinate reaping, threshing and winnowing activities for crops such as wheat, barley, oats, rye, maize, soybeans, linseed, and oilseed. They separate the less nutritional straw from the grains on-site, allowing the farmer to leave the straw as a crop residue on the field or use it to feed livestock . Modern machines are equipped with communications technology that allows the various machines working in a field to automatically synchronise their operations to avoid duplication of activity or double application when spraying, fertilising or seeding (Figure 1) .
Figure 1: John Deere combine harvester coordinating with other vehicles using MachineSync technology. This technology supports the unloading of the combine on the trailer to avoid spilling of grain. © John Deere | Source
3. On-line yield monitoring and mapping
Combine harvesters can also be equipped with yield monitoring sensors that capture the composition and condition of the harvested grains in real time to produce yield maps . When connected to the machine’s Global Navigation Satellite System (GNSS), (ie GPS, GALILEO, GLONASS or BEIDOU), yield maps can identify low and high performing areas of the field . By monitoring yield histories year-to-year, areas with consistently low yields can be identified, enabling farmers to vary the application of agrochemicals appropriately  . There is an example of a yield map in Figure 2.
Figure 2: An example of a yield map. Orange and red zones show low yield areas Source. Generated by Ag Leader SMS software.
4. Automation and robotic assistance
When it comes to harvesting, it is important to ensure that the harvester follows the same route as the planter, to avoid driving over the crop or compacting the soil. Satellite-based positioning systems that modern tractors are equipped with, enable drivers to determine their position to within a couple of centimetres   so they can reproduce the same route for different purposes, such as sowing, fertilizing and harvesting. Auto-steer systems (video hosted on YouTube) are also available, some of which can perform simple turns, but a driver is still needed to monitor the equipment and to watch for obstacles . Robotic systems are being developed though, so autonomous harvesting and precision processes will soon be possible .
5. Advanced modelling and scenario planning
Machine learning (video hosted on YouTube) is increasingly being used in precision agriculture to associate yield outcomes with different properties of the weather, soil or plants. Examples of how machine learning can be used in precision agriculture include:
- Irrigation and water management 
- Predictive crop models to improve returns through rotation and optimal use of land 
- Weather forecasting adapted to farm operations 
- Disease diagnosis .
The benefit of this technology is that it ‘learns’, so can adapt its operations over time and make the calibrations necessary for providing appropriate suggestions to the farm manager.
Lower cost solutions
Many of the Learners in the first run of this course raised the very important point about lower cost solutions. The technologies described above are very expensive. How can famers in developing countries and smallholders afford to implement them? One example of a project addressing this problem in Africa is Hello Tractor, that connects tractor owners and smallholder farmers in Sub-Saharan Africa through a farm equipment sharing application. Cloud Computing technologies provide opportunities for farming communities to share crop data and sensing information. Can you think of ways in which precision agriculture practices could be adapted and applied in low cost contexts?
There are some additional examples of low cost technologies under the ‘See Also’ heading below.
References can be found under the ‘Downloads’ heading at the bottom of this Step.
© EIT Food