Smart Solutions

Drones in Agriculture

The use of drones for precision agriculture, efficient application of weed control or fertilisers and optimised field management.

Implemented inMarathon

Country : Greece

What’s the solution?

One of the newer developments in digital technology is the increasing use of small unmanned aerial vehicles (UAVs), commonly referred to as drones, in agriculture. Drones are remote-controlled flying machines without a human pilot on board.

Drones come in different types, including vertical-take-off and landing (VTOL), multi-rotor drones - particularly quadcopter & octocopter drones with four and eight rotors respectively. These fly more like a helicopter, with the ability to change direction in all directions, but can be more noisy. Other types are fixed wing drones which need to be launched and fly more like an airplane or glider, are less noisy but turn more slowly. Different types of drone are better adapted to certain tasks more than others (see below).

Drones can be fitted with cameras, sensors and even spraying equipment according to their intended use. They have enormous potential in agriculture to support evidence-based planning and spatial data collection.

Drones have two fundamental categories of application in agriculture. They can be an “eye-in-the-sky”, providing optical observation and recording of a territory including crop growth, weeds, presence of standing water and livestock monitoring etc. They can also be a “hand-in-the-sky” with potential applications including spraying (e.g. of fertilisers or weed-killers) and goods delivery across the territory.

The specific type of drone and attached equipment needed depends on the intended use:

  • Crop monitoring: 1) Multi-rotor drone 2) Thermal camera – thermal imagery 3) RGB camera – RGB imagery 4) Hyperspectral camera - hyperspectral imagery (of selected spectra for specific disease) 5) Lidar sensor - point cloud data 6) IoT weather station - environmental and soil data.
  • Drone spraying: 1) Meteorological IoT stations - weather data during sprayings 2) Spraying drone and spraying components (i.e., various nozzle types, computer vision systems) - spray drift, deposition and coverage, and canopy penetration samples 3) Conventional ground sprayer - reference data for common practices.
  • Livestock monitoring: 1) Multi-purpose drones with RGB and thermal cameras - georeferenced RGB and thermal imagery; 2) GPS collars - ground-truth data of the animal locations; 3) Aptimiz sensor - labour time recordings of different farm tasks.
  • Forestry and biodiversity monitoring: 1) Fixed-wing drone with high autonomy; 2) Multispectral and thermal cameras – georeferenced drone imagery; 3) On-board edge device, for data pre-processing and communication handling; 4) Satellite imagery – Sentinel-2 data.
  • Rural logistics: 1) Vertical take-off and landing (VTOL), quadcopter and octocopter drones; 2) Photogrammetry equipment for georeferenced 3D model and navigation; 3) Meteorological stations for each of the points of interest - environmental data; 4) Server for the DD-FMS hosting; 5) Custom-made cargo systems.

These tools and technologies can be used as part of an integrated approach to precision agriculture, together with a range of air and soil sensors, which all together feed data into a centralised database and decision-making system for informing watering, application of pesticides and the optimum time for harvesting.

Although drones are becoming more popular and affordable, they are still considered a costly investment. Other societal issues such as knowledge gaps, data protection and regulatory restrictions, as well as environmental conditions, limit the use of drones. However, research initiatives such as the European ICAREUS project are seeking to improve understanding of the risks and added value of drones in agriculture and scale up their use.

What makes it smart?

The solution uses modern technologies - in the form of drones and IoT devices - to improve a traditional industry - in this case broccoli production – increasing yields and product quality.

  • The use of digital technologies in primary production optimises production and reduces costs through evidence-based decision-making, helping to ensure the long-term viability of economic activity and safeguard jobs in the region.
  • The use of drones also strengthens the capacity for more sustainable forms of agriculture and forestry, reducing the use of water, fertilisers and other resources in the agricultural process – by increasing the available information to be able to provide only what is needed.
  • The solution can be applied in an almost unlimited set of local contexts and uses, including for monitoring crops and plant health, livestock and animal health, forestry including tree health, as well as other forms of mixed land use.

How is the solution implemented?

  • Define the local needs and goals (Use Case Plan)
  • Determine the tools that are critical to meeting the above goal.
  • Define the trial areas
  • Purchase/rent of equipment
  • Consider combining the use of drones with on-the-ground sensors as part of an integrated approach to precision agriculture.
  • Data collection in the field
  • Pre-processing and analysis of data
  • Mapping and reporting
  • Support for agricultural decisions (irrigation, fertilisation, harvesting, etc.)
  • Replication in each growing season

In what local context has it been applied?

Marathon is a town in Greece located on the outskirts of Athens in the regional unit of East Attica. It is known in Greek history as the site of the Battle of Marathon in 490 BC, in which a Greek herald was sent from Marathon to Athens to announce the victory, an event which created the marathon race. Today, the town is a popular holiday resort and agricultural centre – it is the most important domestic sources of vegetable production for the Athenian market.

Agricultural producers are continually under tremendous pressure – from consumers and legislators - to reduce their costs and their environmental impact.  One important means of addressing both of these challenges is to increase efficiency in the use of resources – particularly the use of water, fertilisers, machinery and labour. This is particularly the case in Mediterranean countries such as Greece, where climate change is only expected to increase the pressure of water resources.

Reducing consumption of resources in agriculture requires a solid understanding of the real-time needs of crops and vegetables. The use of sensors and remote monitoring as part of a precision agriculture approach can enable this.

Who was behind the implementation?

Agricultural University of Athens (AUA) led the pilot project to use drones in agriculture in Marathon

NEUROPUBLIC (NP) is a private technology provider which was a key partner of AUA providing, for example, the required IoT devices, cloud infrastructure and decision-support services.

Marathon Bio Products was the key local agricultural partner growing more than 20 vegetable varieties throughout the year.

What was the local journey?

  • The decision was taken to test the use of drones in agriculture in the Marathon area through a project led by the Agricultural University of Athens, together with the technology provider NP and local producers ‘Marathon Bio Products’.
  • Through discussions with local producers, a detailed plan and experimental protocol were established by the AUA, focusing on the production of organic broccoli. This plan included definition of the exact areas of operation, methods, technical requirements, equipment to be used for data collection, and calendar of activities.
  • The boundaries of the fields were geo-referenced using GPS technology, as soon as the experimental fields were chosen. Management zones were formed using soil electrical conductivity (ECa) mapping, assisted by elevation mapping using Centimetre-grade Real Time Kinematic (RTK) GPS.
  • The technology company NP provided a multi-rotor drone equipped with a multi-spectral camera as well as IoT devices which were installed in the end user’s field, including a fully operational network of IoT stations (for meteorological and soil data) as well as proximal sensors - multispectral canopy sensors and chlorophyll metres. NP also retrieved satellite data from Sentinel 1 and 2 for the target fields - accessible free of cost from the Copernicus Open Access Hub.
  • In addition, NP provided the cloud infrastructure to manage and process the collected data, as well as decision-support services – through its Gaiasense service -to issue advice and risk warnings on irrigation, fertilisation and pest control.
  • Drone flights (Phantom 4 Pro) were conducted every fifteen days and images were captured with a multispectral UAV camera for vegetation index calculation and map generation
  • The data collected by the drone and the other sensors was used to identify weed patches and determine the maturity of vegetables for optimal harvesting and yield estimation.
  • Based on this information, robust predictive models for yield and quality of harvested product were developed and provided to growers for decision making in subsequent years.

What have been the main outputs & results?

  • Development of a yield prediction model for organic broccoli based on multispectral UAV imagery with over 84% accuracy.
  • Optimisation of irrigation, fertilisation and pest management models for organic broccoli with reduction of inputs (water, nitrogen and organic pesticides) by over 25%, 38% and 61% respectively.
  • Increase the percentage of "A" grade quality produce by 15% and total yield by over 11%.
  • Achieve over 90% accuracy in weed detection through drone imagery at the early development stage of broccoli.
  • Integrate and evaluate a real-time automated warning system for broccoli cultivation.

What does it bring the village/community?

By increasing the productivity and quality in the cultivation of outdoor crops and vegetables, the use of drones (within an overall approach to precision agriculture) can improve the long-term viability of local agricultural production and the competitiveness of local farmers – protecting and creating jobs and economic opportunities locally.

The most important contribution is to the social fabric of rural communities, where increased farmer income and the reputation of better crop quality can have a significant impact on the community and increase local prosperity and quality of life.

By increasing the resource efficiency of agricultural production, the solution can also make a contribution to mitigating the impacts of climate change and reducing negative environmental impacts from agriculture.

What’s needed

Financial resources

Main types of cost:
Financial needs:

Set up / Investment costs: EUR 40,000

• Drone and camera

• Sensors

Other sensors and the data management platform were provided in kind by AUA.

The initial cost for the drone equipment is around 20,000-30,000 euros (drone platforms and payload -sensors)

Ongoing costs: the approximate total cost of applying drones and other digital equipment in agriculture and providing it as a service to the end user, from data collection to final reporting, is EUR 10,000 per ha per year, including maintenance, personnel, travel costs, etc.

Funding received:
Horizon 2020 research and innovation programme: SmartAgriHubs project40,000 €Drone, camera, sensors and piloting actions.
AUAContribution in kind: data platform and additional sensors

Human resources

• A certified drone pilot

• Agronomists and/or farmers

• Smart agriculture experts and advisors

Physical resources

• A multi-rotor drone (unmanned aerial vehicle) with multispectral camera

• IoT devices including stations for meteorological and soil data collection.

• Proximal sensors - multispectral canopy sensors and chlorophyll metres.

• Cloud infrastructure to manage and process the collected data, as well as decision-support services

What to do…

  • Determine the producer's individual goal and the needs to be met.
  • Determine the tools that are critical to meeting the above requirements.
  • Determine the causes and extent of variability in the field.
  • Consult experts to learn about software systems and machine capabilities to decide which tool is right for you.
  • Know the local legislation on drone use.
  • Register the drone and receive a certification as a drone pilot.
  • Take care of underperforming areas.
  • Create soil conductivity maps, elevation maps and yield maps.

and not to do

  • Do not be impatient. Successful implementation of drones in agriculture, forestry and rural areas needs time and knowledge enhancement. It may not show immediate results.
  • Do not be discouraged. It may take a while to get used to a completely new technology in agriculture. Drones are weather-sensitive: you may lose or even crash a drone or the quality of images may be spoiled.
  • Do not hesitate to seek advice and explore the many existing possibilities and technologies of precision agriculture.

Funded by