Agricultural robots are data acquisition tools at heart

Agrochemicals and old-school machines capture nearly the entire value in global crop production. A mega shift however is well underway that will see data acquisition and analytics take an ever large share of the overall value. Agricultural robots will accelerate this enormous global value chain reconfiguration, bringing the endgame of ultra-precision automated farming ever closer.

 

Agricultural robotics are already a market reality in parts of agriculture. Indeed, the IDTechEx report Agricultural Robots and Drones 2016-2026: Technologies, Markets, Players finds that agricultural robotics is already a $3bn market and will reach $10bn by as early as 2026. In this time IDTechEx Research finds that robots will enter ever more aspects of agriculture and that robots will transfigure from static, repetitive machines into mobile and intelligence ones. Read the report to learn how this mega transformation will take place.

The concept of precision agriculture has been around since 1980s but its uptake has accelerated since the mid-2000s. This is partly because key enabling technologies such as RTK GPS have fallen in cost and partly because the entire ecosystem has started to come together.

 

 

Data is already becoming important in agriculture. For example, farmers employ aerial data maps of their fields to optimize the application of nutrients. Variable-rate technology (VRT) with GPS-enabled equipment is already an incremental improvement over the broadcast application of inputs across the entire farm, uniformly sprayed regardless of the needs of specific patches.

 

VRT technology is however only a stepping stone towards the endgame of ultra-precision agriculture. Here, farm management evolves past VRT towards being administered on an individual plant basis. This means that inputs are dosed according to exact needs of individual plants. Robotics will be the enabling technology for this ultra-precision agriculture.

 

 

Robots are already accelerating the evolution of farming practises towards ultra-precision agricultural. Autonomous drones equipped with multi-spectral sensors are flying over farms, providing indicative information about the health and status of crops.

 

 

These drones will lower the cost of cm-level high-resolution images, particularly for smaller farm owners. In time, they will render more accurate data available to more farmers compared with satellite or plane techniques.

 

The drone hardware will soon become increasingly commoditized with the value shifting to data analytics or specialized sensor hardware. The algorithms will improve thanks to the availability of more data, cloud computing and the engagement of specialized agro-scientists. Soon, crop disease- or condition-specific software will be available via dedicated app stores, helping farmers receive actionable advice and not just maps or raw data.

 

In the meantime, the sensing hardware will also become highly specialised, converting expensive data-rich hyperspectral sensors into inexpensive specialized purpose-built multispectral ones. This, in turn, makes aerial spectrometry available to more farmers, increasing the overall volume of farm data.

 

Overall, the IDTechEx report Agricultural Robots and Drones 2016-2026: Technologies, Markets, Players forecasts that agricultural drones will become a $485m market in 2026.

Progress is not just limited to developments in the air. In fact, improved perception and artificial intelligence are already enabling precision weeding on ground: organic farming has already adopted robotic weeding implements whilst deep learning has been brought to tasks such as precision weeding or lettuce thinning.

 

 

In its simplest form, a template-matching or crop-following algorithm together with an RGB/NIR camera identifies weeds as out-of-row living objects and adjust the position of a mechanical hoeing implement to eliminate them.

 

In more advanced forms, millions of images of weeds and crops are being used to train the detection algorithms, enabling the robots to identify greater varieties of crops and weeds in less structured, more complex, and changing farm environments.

 

This is still in its early days but the direction of development is clear: deep learning based on huge troves of data will be increasingly applied to agriculture. This means that companies must seek to carve out a strategic 'data' position for themselves in this emerging value chain and must become more competent in managing and processing large amounts of data.

 

IDTechEx Research forecasts that the market for de-weeding and thinning robotic machines and implements will exceed $620m in 2026. For more granular data refer to the IDTechEx Research report Agricultural Robots and Drones 2016-2026: Technologies, Markets, Players.

In parallel, autonomous navigation has become technologically possible. Indeed, autosteer is also commonplace in agriculture with more than 320 k tractors with autosteer or tractor guidance capabilities sold in 2016. This figure is projected to rise to 620 k/year by 2026. This makes tractors the earliest and largest adopter of autonomous navigation technology globally.

 

 

This trend will continue. The autonomous navigation technology will become lower cost meaning that more sensors will be integrated into mobile robots. This, in turn, is facilitating an evolution from manned autosteer technology towards safe, unmanned, and autonomous navigation. This can profoundly impact the way agriculture machines are envisioned by taking the driver out of the equation. It will upend the commonly-held notion that big is better.

Ultra-precision agricultural practices will come as a natural evolution of existing trends. This is because farmers' conservatism together with the slow pace of change limited by harvest cycles turns many potentially revolutionary robotics technologies into evolutionary ones.

 

The drone deployment will become accelerated in the coming years as regulatory barriers fall and the hardware becomes commercialized. The challenge will remain in developing analytics that deliver actionable advice and not just raw data maps, otherwise farmers will see this trend as just a fad.

 

Agricultural ground robots will first find traction as robotic tractor-implements or small autonomous vehicles aimed at small farms. Many will first be offered as a Robotics-as-a-Service (RaaS) priced in $/acre or $/Kg.

 

This business model will be commonplace in the early years because it reduces the upfront expenditure and because it compensates for the lack of farmers' trust and the absence of total technical reliability by having a skilled operator.

 

 

Standard equipment sale business, however will grow more dominant towards the end of the decade as suppliers try to geographically scale and rely on the existing network of dealerships to sell their products.

 

To learn more about agricultural robots and drones and the future of farming read the IDTechEx report "Agricultural Robots and Drones 2016-2026: Technologies, Markets, Players" at http://www.IDTechEx.com/agri. Here, a detailed technology roadmap outlining how robotic technology will enter into different aspects of agriculture, detailed territory-segmented ten-year forecasts showing the market future of different agricultural robotic and drone technologies, and detailed interview-based profiles of leading companies.