From new sensor technology to the use of artificial intelligence, agriculture is embracing smart technologies. By Neil Tyler.

Innovative technology is being widely deployed and embraced by the agricultural sector.

Whether it’s the use of monitoring drones that are capable of analysing the use of farmland, the deployment of sensors that can monitor soil conditions, unmanned farming vehicles that harvest crops, or the use of artificial technology (AI) to better understand changing weather conditions the way in which the farm industry operates is changing – in fact the term, Agriculture 4.0 in now being used to represent the changes being seen in farming and the way in which these technologies are being implemented.

Advances such as automatic driving of agricultural machinery and the introduction of robots and drones are expanding the scale of production seen on large-scale farms and smarter agriculture is seen as necessary to deliver higher-quality production and larger harvests. Technology also has a vital role to play in terms of saving on labour costs and freeing up workers from having to undertake difficult and, at times, dangerous work, while at the same time enabling non-experts and relatively inexperienced workers to handle multiple different forms of work.

A shortage of labour in the agricultural space has meant that a growing number of companies and start-ups are developing and manufacturing agricultural robots to assist farmers with numerous farm operations, whether that’s fruit-picking, harvesting, planting, transplanting, spraying, seeding, or weeding. Used to automate repetitive tasks smart agricultural machines such as GPS-enabled autonomous and semi-autonomous tractors are being used for harvesting, while robots are also automating systems in livestock management.

Robots that are equipped with stereo cameras can now identify and pick fresh produce based on their size and ripeness and can even make informed decisions based on data that’s collected and processed.

By using AI farmers can incorporate and use real-time insights from their fields, allowing them to be more proactive and make better and more informed decisions. AI and machine learning (ML) are also being used in disease recognition in plants and livestock making it possible to detect and deliver corrective responses.

Weather prediction

Weather forecasting has an important role to play in assisting farmers to deliver both healthier and more significant crop yields. Estimated precipitations and temperatures alongside historical data are used to determine seeding or harvesting, and can determine when and how fertiliser and herbicidal treatments are used - some chemicals need to be sprayed on a dry day, for example, while others require moisture in order to be activated

So reliable agriculture weather reports can not only help to boost harvests but can save farmers money and reduce risk.

Climavision, a US-based weather tech start-up, has created a new artificial intelligence (AI)-powered weather radar and satellite network that combines lower altitude, proprietary data with machine learning and AI technology.

According to Chris Goode, the company’s CEO, “Our sensing network was developed to fill gaps in the coverage that’s provided by the existing NOAA and NWS systems across the US and while the current weather surveillance model provides a picture of weather at a given moment, the picture is not complete.”

At present institutions and businesses around the world rely on a network of weather radars that look to collect weather data, in real-time. “Weather balloons and aircraft sensors are also deployed to collect additional data points at different parts in the atmosphere,” explains Goode, “but even so there are gaps in coverage especially covering the atmosphere where volatile weather forms, and at the lowest levels, where this weather occurs.“

According to Goode the company’s system fills in those gaps with critical weather radar infrastructure that brings together real-time data with space-based observations to get a complete view of what’s happening from the ground up.

“Once the date is collected, we use AI, machine learning and IoT to process, interpret and distribute the data,” says Goode.

Traditional weather observations consist of multiple variables, including temperature, wind, pressure, and humidity, which are interconnected and complex and Climavision uses AI to carry out quality control on those variables with the data having been collected from different platforms and various channels.

“We use AI to extract the most impactful meteorological information and to integrate the information from different platforms,” according to Goode.

“By providing a better understanding of what’s happening in the atmosphere, whether that’s hours or even just minutes earlier than would have been the case, there is the potential to save farmers money by knowing what’s heading their way, whether that’s flooding, snow, or tornadoes. It means that they can at the very least be better prepared.”

Sensor technology

According to Juli Ban, Product Manager (Function Devices), Murata, worries as to whether the agricultural industry will be able to meet the demand for food as well as having to contend with different environmental factors means that, “Technology will need to play a pivotal role in next-generation agricultural activities. Technology will enable the crop yields per hectare of land to be significantly increased, and make better use of essential resources, while also keeping labour costs down.”

With the United Nations (UN) expecting the global population to top 10 billion within the next 35 years, as well as environmental degradation prompting calls for more sustainable solutions across all industries, the pressure on the sector is immense.

“Smart agricultural practices will be critical in managing these competing pressures in terms of the acquisition and subsequent analysis of data,” said Ban. “More sensor devices will be need across farming sites to provide continuous updates on key parameters affecting food production (such as air temperature, humidity, degrees of illumination, etc.). This will mean that farmers will be able to gain the information they need to respond to changing circumstances - whether that is short-term fluctuations in growing conditions or the uncovering of trends which will need a longer-term solution.

“The sensors used in an agricultural context need to have certain characteristics if they are to be effective in their assigned function. Firstly, they must be robust and, secondly, they need to deliver accurate data - otherwise the decisions made by farmers may be incorrect, with crop production volumes or crop quality suffering as a result. Finally, these sensors have to be accompanied by appropriate connectivity.”

Constantly updated parametric data in relation to soil quality is critical, allowing farmers to gauge the amount of nutrients and salt ions present, and checking for sufficient rainfall, gaining insights into the quantities of fertiliser required or looking for signs of groundwater pollution.

“The distribution of soil sensors will depend on the nature of the crop being grown, and how closely parameters need to be monitored. For higher value crops greater sensor density will be justified. How regularly data needs to be acquired will also be crop dependent, as well as the actual methods being used,” said Ban.

Soil samples will basically consist of the soil mineral itself (in the form of grains), air gaps and pore water. In-depth studying of soil conditions will require electrical conductivity (EC) measurements to be conducted and this will give data on the composition of the soil based on the resistance properties of each of its constituent parts.

“One of the problems with existing soil sensor devices is that their accuracy can be detrimentally influenced by temperature variations and excessive water content, as well as the presence of chemical compounds. Another problem that needs to be factored in is that rocks between the sensor’s electrodes can interfere with the results obtained. For this reason, a multi-electrode arrangement will prove to be much more accurate,” said Ban.

To address this, Murata has developed an integrated three-in-one sensor device that can determine the EC and measure the electrical permittivity, so as to give a value for the volumetric water content (VWC) figure as well as a temperature sensor element.

“This sensor enables multiple measuring patterns to be generated, thereby eliminating uncertainty about the results.”

According to Ban, “Engineering innovations, like this three-in-one soil sensor, have the potential to make a substantial contribution to farming becoming more data-driven, leading to productivity being increased and running costs lowered.”

Upgrading electrical equipment

“What we’re seeing in agriculture is a paradigm shift which means that, over the next five years, farms will make increasing use of everything from drones and field sensors to autonomous harvesting equipment and cab-mounted GPS units,” says Tom Borland, UK and Ireland Country Manager at PEI-Genesis, a cables and connectors specialist.

“With this change, farmers must update some of their electrical equipment. One component that’s easy to overlook in this process is the connector. Connectors are already used widely across the agricultural sector and are responsible for delivering power, connecting cameras to screens in the tractor cab, and allowing operators to send control signals to harvesting attachments.”

While the connectors used in farming equipment are already designed to handle a variety of harsh conditions, the connectors that are used on farms will need to be fit for purpose going forward.

According to Borland, while stainless-steel connectors are useful for heat shielding, their weight may make them less favourable for drone applications and while plastic or composite connectors offer a low weight and strong connection, they can become brittle with prolonged outdoor exposure and the heat from the sun.

“Farmers must consider the ability to transfer power and high-speed data effectively and this may require high bandwidth fibre optic cables that can collect and process the data from all the sensors as well as the wireless control signals from the remote operator.”

It will become vital in the coming years that farmers choose the correct connector across their smart farm, for everything from lighting and power systems, and steering and motion, to monitoring and control, data signals, powertrain and cabin interfaces.

“At the most basic level, these connectors will be variations of existing cylindrical style M12 ethernet connectors, while at the more advanced end of the spectrum, connectors will make use of the ISO-Bus platform, a system that allows farmers to use different equipment from different manufacturers with the same tractor and towing vehicles, increasing compatibility as a result.”

As the transition to Agriculture 4.0 becomes a reality smart agriculture is expected to spread further into cultivation it will not be possible without connectivity technologies like 5G, LPWAN, rural broadband, or satellite-enabled connectivity.

These forms of connectivity will enable different IoT devices, robots, and sensors to communicate the data at ultrafast speeds and this will enable farmers to monitor the data collected more accurately in real-time and take the required actions.

The high-tech, connected farm is fast becoming a reality.