The introduction of Global Navigation Satellite Systems (GNSS) in agriculture started a new era of farming practices. Thanks to these navigation systems, farmers are able to optimise their cultivation practices and their field traffic. Many futuristic and maybe not so futuristic machinery is becoming feasible, such as crop care robots, autonomous tractors, plant specific treatments and master-slave guidance in the field. But even without these engineering innovations farmers can already do a lot with GNSS. It starts with measuring the exact boundaries of a field and adjust e.g. the sowing to it. Knowing the exact working width of the harvester machine, farmers can optimise the sown area to make sure, for instance, that harvesters always works at a full width of crop to harvest. This saves time and fuel during harvest but also during all other cultivation practices.
John Deere’s Ag Management Solutions deliver faster field performance, reduced input costs and stronger yields
Optimising this field traffic to the cultivated area is called the drive map or traffic plan. Besides a more optimal layout of the field, it also helps in reducing overlaps and gaps in the cultivation. Altogether, these GNSS based optimisations cut costs by 15% on an average, due to the use of less fertiliser and other inputs and less fuel consumption at every field operation.
To maximise this cost reduction, farmers must invest in GNSS-ready tractors involving a different steering system, a GNSS receiver and a board computer/controller. Many farmers prefer to work with the highest quality available GNSS navigation systems. The 15% costs optimisation can only pay for all these investments above certain farm sizes and types. Therefore, besides farmers, contractors to whom field activities are outsourced are among the early adopters as they make more use of their tractors and machinery.
High uptake of GNSS
A recent survey in the Netherlands done by the UNIFARM project shows that even at smaller farms GNSS is now penetrating well, although larger farms clearly have a higher uptake rate. In the Netherlands, 65% of the arable farmers use GNSS in their cultivation. And although precision farming mostly talks about location-specific treatments, just optimising the geospatial aspects is the most popular entry technology into precision farming for many farmers in the Netherlands.
Interestingly, in USA, Canada and Australia — the birthplace of precision agriculture — due to their large fields with homogeneous crops, the use of yield monitors is often seen as the main entry technology for precision farming. Spatial differences in crop yields reflect differences in soil conditions that can, for instance, be repaired with fertiliser. These conditions are structural and often reflected in the same patterns in yields. But the role of yield monitors is very low in the Netherlands as compared to those coun- tries. Here the GNSS is much stronger.
The increase in GIS tools on farm to make these optimised field plans also help farmers to look at spatially varying yields. After being able to optimise their traffic, these spatial variations also become more relevant. Linking the final yield to all the operations done in the season provides an excellent benchmark to improve the farm performance.
Having said this, precision agriculture is not only doing the right thing at the right place and moment, but it is also documenting these activities for analyses purposes. This will be a growing field now that geospatial technologies have become common practice. Although difficult to express in percentages of costs reduction, the automatic documentation, thanks to precision farming tools is a highly valued ‘bonus’. Also machine manufacturers discover this as a relevant add-on to their machines and develop farm information systems that seamlessly integrate with their machinery and equipment.
Large machine manufacturers as John Deere and Claas have these systems in place, just as GNSS systems provider Trimble has. The latter has already taken the next step towards sensoring systems that are needed to provide the information for creating task maps. The Trimble portfolio now includes a tractor-mounted sensor and a Remotely Piloted Aircraft System (RPAS), that can fly over crops and fields and with a right camera collect relevant remote sensing imagery that help farmers make decisions about when and where and how much fertiliser to apply — among others. In this way, these machine and equipment manufacturers lead the way to make field work more location specific. This requires integration with all kinds of GIS data like soil maps, groundwater data, remote sensing maps of the crop etc. There is an explosion of possibilities from scanners, sensors, sampling strategies, cameras and more. And the interface between sensors and cultivation practice is: the task map — the spatial instruction set indicating the different application rates at different zones and spots in the field. Applications that are already demonstrated in practice and have created high expectations are, for instance, variable sowing densities based on soil heterogeneity (and field geometry) and variable fertilising based on satellite imagery.
Farming is simple, but not easy
At the other end of the spectrum, tractor mounted multispectral cameras control the flow of fertiliser or other crop chemicals to provide the right dose at the right spot. This is often called near-by sensing — as opposed to remote sensing done by satellites or airborne sensors. For many farmers, this is an ideal solution as there is no human intervention needed to translate sensor readings into a task map. This increases significantly the usability of the tools. However these black box-like solutions have limited ‘knowledge’ of the causes of different sensor readings, therefore not easy to interpret.
As often seen in the application of innovative technologies, developments start from a single viewpoint and thus we have seen information systems make task maps fully dependent on satellite data and simultaneously information systems around real-time sensing with on-board sensors and processing. The combination of different sources of data is not always seen as a logical pathway, although everybody realises that only a combination of factors, often measured in different ways and at different time and spatial scales, results in the best possible solution. Now here is the catch: this requires integrated information systems and decision support tools for farmers to weigh the different sources and qualities and to evaluate different scenarios.
Tackling spatial variation
There are three possible routes to deal with spatial variation. The first one is to make robust management zones based on available spatial knowledge like soil maps, previous year(s)- yield maps or other sources indicating spatial heterogeneity. These management zones can already make relevant differences in application rates when directly translated into task maps. They can also be used to provide context to in-season sensing, in particular to differentiate two spots with similar crop index values but with different history, requiring different responses. All previous applications or cultivation practices and previous sensing maps can be considered as available spatial knowledge and can contribute to develop (or adjust) these management zones.
The second and third routes involve in-season sensing to measure the crop response to all variability it is dealing with. The most noticeable difference between remote sensing and near by sensing is the ability to get processed in real time and make a variable rate application based on sensor-readings and pre-programmed conversion rules. The advantage is that it does not require human interference and precision farming can be done without additional education or training. It just works, and it uses the latest information from the field. As an alternative, sensor data can be taken into the office and can be combined with farmer’s knowledge and management strategies.
A clear characteristic of the in-office processing is that it takes place ‘off line’. It has the advantage to take on more knowledge and inputs. Farmers can even compare different scenarios and thus make optimised decisions. And it can be used to determine the timing of application: Maybe it is better to wait a day or two. The disadvantage is of course that it introduces an extra task to farm management which often means that after a long day of field work, farmers need to sit behind their computer. This situation has not really evolved in standing practices, but it can be expected that it will soon lead to new and better farm management software with more GIS capabilities.
The high expectations from unmanned aircrafts, or remotely piloted aircraft systems (RPAS), lay in the fact that they are available to farmers at a much shorter ‘delay’ than satellite imagery or (traditional) aerial photography. It can be deployed when needed and processed within an hour, so in terms of sensing it is almost real time. It fits very well in the new way of farming where farmers optimise their inputs and labour in cultivation practices. This can only be done with adequate information.
The Netherlands is internationally always characterised as a country with small farms with small fields. These small fields and the long history that farmers have with their land is a hurdle to take with innovative technology. Farmers in the Netherlands however show, that although size does matter, smaller farms benefit from GNSS. As a farmer once put it, GNSS technology allows him to do large scale agriculture in a small scale landscape. For farmers who adopted GNSS already and “know where they are” in the field, the variable rate technology becomes their next step in precision farming. With the help of smart technology and GIS, farmers will be able to optimise their activities. In the end, it all comes down to the ability to deal with complexity. Farming is simple, but it is not easy.