Space-enabled Crop disEase maNagement sErvice via Crop sprAying Drones (SCENE-CAD)

Lead Research Organisation: Loughborough University
Department Name: Aeronautical and Automotive Engineering

Abstract

Crops diseases are widely considered as one of the main challenges in modern intensive agriculture. They not only damage crop yields and quality, posing serious threats on food security, but also exert adverse impact on the environment due to inappropriate and ineffective treatment of using excessive pesticides. The overarching goal of this project is to develop and deliver a turnkey crop disease management service to key stakeholders in China and other relevant countries, which will support them in early detection, rapid response and targeted intervention of major crop diseases. This goal will be achieved by combining a number of crop disease monitoring and forecasting technologies developed from the previous STFC UK-China Newton Agritech projects and extending their impact through integration with crop-spraying drones, so that the benefits of early identification and forecasting of crop diseases can be consolidated by rapid, targeted, and automated pesticide spraying actions.

This project is structured under six main work packages (WPs), including four product/service development work packages to develop and transfer space-enabled technologies into agriculture services/tools and two demonstration work packages to encourage the acceptance of the developed services and disseminate the technologies to other countries. To facilitate the technology development/transfer, recent advances in image analysis will be assembled and utilised to extract key information from remote sensing data (WP1). Following the cycle of a dynamic crop disease management process, the project will first develop a cloud-based online service to identify, monitor and forecast hot-spot regions of overwintering wheat rusts using crop disease models, information from satellite and drone remote sensing and environmental parameters (WP2). Such information reported by the online service can be used to inform the planning and deployment of crop spraying drones to promptly control the diseases. Second, to guarantee the quality and efficacy of the pesticide delivery, an intelligent and user-friendly planning and management software for crop spraying drones will be developed (WP4), where a parametric drone spraying model (WP3) will be established to characterise spraying deposit distribution and further software tool to assess the spraying quality on different crops/diseases.

The practical benefits and long-term impact of the developed products/services will be demonstrated, with the strong support from project partners in China, through two demonstration campaigns designed in this project. The first one (WP5) focuses on the overwintering wheat rusts in Gansu Province, which is the origin of inoculum that causing yield losses in the main wheat production in the Central China. It is expected that the hot-spot areas of rusts can be effectively identified using the developed service and treated using spraying drones in autumn, thus preventing or reducing the rust epidemics in spring in other regions of China. The second campaign (WP6) is dedicated to showcase the benefits in the case of rapid response to unexpected disease outbreak. The spraying missions can be automatically generated using the developed software based on the disease severity and distribution in a variable-rate manner to ensure the spraying quality while reducing operator's workload and the use of chemical pesticides and fertilisers.

It is envisaged that this project will provide an integrated Agri-tech service for crop disease management that is able to improve the food productivity, reduce both the labour and pesticide costs in practice, and contribute to the long-term sustainable growth in agriculture. Moreover, through the impact activities, the project will have a profound and long-lasting impact on the local crop protection organisations, spraying service providers, drone operators and eventually the farmers, in China and beyond.

Planned Impact

By promoting food security and environment protection through developing and delivering the informative and automated disease management service enabled by space/drone technology, this project will ultimately improve the livelihoods of targeted agricultural stakeholders such as farmers, local communities, plant protection service providers, local and central government and agriculture companies.

In this research, wheat yellow rust will be used as the main example to demonstrate the impact and promote the technological approach. Wheat, along with rice and maize, is one of the three most important staple crops in China and is also the mostly widely grown crop (215m ha) feeding 2.5 billion people worldwide. Alarmingly, 88% of the world's wheat production is susceptible to yellow rust infection, resulting in a yearly loss of £0.8 billion worth of crops. This study, combing modelling and prediction of rust disease with near real-time remote sensing information and automated drone spraying, would facilitate timely and site-specific disease treatment, which not only can preserve wheat productivity, but also lead to a reduced use of chemical pesticides, generating enormous environmental, social and economic benefits to UK, China and beyond.

The substantially reduced use of pesticide due to the newly developed early disease monitoring, forecasting and precision intervention services in this project will significantly mitigate its adverse effects on environment including air (pesticide drift), water (pesticide residues) and soil (damage to the community of microorganisms), and agriculture product (pesticide residues). Therefore, this project will not only help protect the environment but also improve human wellbeing and animal welfare.

Early disease monitoring and timely intervention can maximally preserve crop productivity by using a reduced dose of pesticide, and therefore will also improve profitability on farmlands. This is particularly important when considering the enormous scale of wheat production in China and worldwide. Higher annual incomes due to introduction of the developed services will benefit the health of the rural poor, particularly the Gansu mountainous region, one of the least developed regions in China. This will help to reduce the malnutrition prevalent in the under 5's, and the increased health benefits of the working population and their offspring. Improving workers' income will make paying for medical care and reduce general domestic disharmony. In addition, the adoption of drone spraying instead of manual backpack spraying will significantly reduce the risks of exposure to harmful agents, benefiting equally male and female smallholder farmers.

Beyond the direct impact in agriculture, the space and drone enabled precision management technologies can also find a wide range of other applications in related areas. For example, it can be directly applied to forest protection and monitoring including fire detection and monitoring and pest/disease mapping. It could also be applied to pollution monitoring and tracking, flooding prediction, etc., generating more and wide impact.

This project can promote the long term, sustainable and innovative partnership, which will support knowledge sharing, uptake of agricultural technologies and technology transfer within and between the UK, China and third country partners. New Chinese partners (e.g. GAAS, NAPPA, BAT) and third world country partners in South Asia and Africa will all contribute to personnel and knowledge development, which will facilitate future Agri-tech projects and work to expand the scope and ultimately deliver new social, environment and economic opportunities.

Publications

10 25 50
 
Description Farming Innovation Pathways (FIP) - Integration of UAV with UGV in Agriculture Scenarios
Amount £242,794 (GBP)
Funding ID 10004402 
Organisation Innovate UK 
Sector Public
Country United Kingdom
Start 10/2021 
End 09/2022
 
Description Newton Fund Researcher Links 2020-RLWK12-10654
Amount £24,000 (GBP)
Funding ID 2020-RLWK12-10654 
Organisation British Council 
Sector Charity/Non Profit
Country United Kingdom
Start 02/2021 
End 12/2021
 
Title Snow Coverage Mapping by Learning from Sentinel-2 Satellite Multispectral Images via Machine Learning 
Description The dataset for snow coverage mapping with three typical classes (snow, cloud and background) is collected and labelled via the semi-automatic classification plugin in QGIS. 
Type Of Material Database/Collection of data 
Year Produced 2022 
Provided To Others? Yes  
Impact This work investigates snow coverage mapping by learning from Sentinel-2 satellite multi-spectral images via machine-learning methods. It is shown that (1) both conventional machine-learning and recent deep-learning methods significantly outperform the existing rule-based Sen2Cor product for snow mapping; (2) U-Net generally outperforms the random forest since both spectral and spatial information is incorporated in U-Net; (3) the best spectral band combination for snow coverage mapping is B2, B11, B4 and B9, even outperforming all spectral band combinations. The snow coverage is useful in many applications including crop disease forecasting. 
URL https://github.com/yiluyucheng/SnowCoverage
 
Description Chinese Academy of Agricultural Sciences 
Organisation Chinese Academy of Agricultural Sciences
Country China 
Sector Academic/University 
PI Contribution The project team provided the expertise in analyzing and interpreting the field test data and contributed to the paper writing.
Collaborator Contribution The partner conducted several field works to collect data for the project. The data collected by the partner are critical in understanding the wheat powdery mildew under different nitrogen input levels.
Impact http://dx.doi.org/10.3390/rs13183753
Start Year 2020
 
Description GAAS 
Organisation Gansu Academy of Agricultural Sciences
Country China 
Sector Public 
PI Contribution We provide satellite remote sensing techniques to estimate snow covering in Gansu area of China to help establish the understanding of snow cover on wheat yellow rusts overwintering phenomena. The overwintering wheat rusts in Gansu Province is the origin of inoculum that causing yield losses in the main wheat production in the Central China. It is expected that the hot-spot areas of rusts can be effectively predicted and identified using the technology developed in this project, thus will reduce the rust epidemics in spring in other regions of China.
Collaborator Contribution GAAS helped us to identify 20 sites of wheat fields in northwest of China to collect wheat rust data for wheat rust modelling purpose.
Impact Dataset for wheat yellow rust in overwintering phase.
Start Year 2020
 
Description NAFU 
Organisation North West Agriculture and Forestry University
Country China 
Sector Academic/University 
PI Contribution We developed remote sensing technology using satellite data for wheat field classification in mountainous areas and data analysis methods for wheat rust modelling.
Collaborator Contribution Northwest Agriculture and Forestry University (NAFU) helped us identify 20 wheat fields in northwest region of China to investigate yellow rust disease at different levels for overwintering phase. The researchers at NAFU also take the ground truth data manually for the comparison with remote sensing results.
Impact Dataset about wheat rusts during winter period
Start Year 2020
 
Description Newton Fund Researcher Links workshop on Crop Disease Monitoring and Forecasting 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Professional Practitioners
Results and Impact Effective monitoring and management of crop diseases is one of the main challenges in sustainable agriculture. A two-day Researcher Link workshop, aiming to share the knowledge and identify key challenges in applying information and big data technologies to crop disease control, was co-hosted in December 2021 by researchers from Loughborough University in the UK and Northwest Agriculture and Forestry University in China. The workshop was in-person at Loughborough University and online with Chinese partners (about 60 participants in total).

The workshop provided a unique platform for Early Career Researchers (ECRs) from both UK and China to share their research and foster future collaborations in precision agriculture. The outcomes of the workshop include the following aspects:

1) Both recent advances and future challenges in smart agriculture and farming have been discussed in the workshop. In total, 21 multidisciplinary presentations, ranging from crop protection to data processing technologies and agricultural robotics, were delivered in the workshop, majority of which were from ECRs. The discussions and interactions in the workshop will facilitate the participants and their colleagues to focus on the key research questions and generate more collaborations in the future.

2) The workshop provided opportunities for the participants to engage with industry and business by inviting relevant stakeholders. It was featured with representatives from two UK agriculture robotic companies and a UK based research institute who is closely working with farmers and growers. The relevant discussions will help researchers to better understand the industrial needs and in turn provide better technical solutions and agriculture services to end users in both countries.

3) The workshop participants from both countries have gained more contacts and extended their networks during the workshop. Many researchers across different institutes have been able to develop opportunities of following-on discussions, research proposals, and possible academic publications. More potential collaborations in both short and long terms will be expected and supported by the workshop mentors.
Year(s) Of Engagement Activity 2021