Mobile Soil Health
Integrating Visual and Context Information into a Mobile Intelligence Solution for Sustainable Management of Wheat Pests and Soil Health
The technical feasibility of integrating visual and contextual information with advanced data fusion techniques into a mobile pest management solution that offers:
Quantitative Analysis
Rapid detection and effective quantification of wheat pest
Pest Management
Places pest quantification into context of regionally/nationally relevant pest tolerance thresholds
Decision Making
Determines whether a pesticide application is advised
Innovations
Lightweight Pest Quantification Techniques
The solution delivers a new optimised lightweight CNN model for rapid and accurate wheat PEST quantification running in mobile devices. This technique will be able to support many smallholder farms located in remote areas, with inconsistent network coverage.
Robust and Effective Pest Quantification Technique
The solution delivers a new board learning data fusion algorithm CNN model for effectively fusing hybrid and local activities features with contextual information. This technique will be able to achieve high accuracy and good robustness of pest detection and recognition in nature scenes.
Sustainable Pest Management Solution
The solution enables forecasting of accepted threshold of wheat pest and estimation of efficacy of pesticide use after detecting wheat pest. Efficient and sustainable crop protection is of vast economic and ecological significance for food and feed production worldwide.
Project Team
The project team comprises the University of Sheffield (UoS), ADAS, Mutus-Tech (MT).
UoS, MT and ADAS have worked together on previous and ongoing projects (Newton Network + AgriTech, Innovate UK-China Smart Farming).
The Department of Computer Science at UoS will be responsible for designing and optimising the pest recognition and data fusion technique.
The technical investigation of this project will utilise the PestNet model developed in a recent UK-China smart agriculture project and an advanced data fusion technique developed in a previous project.
ADAS is the leading provider of independent agricultural consultancy within the UK. The ADAS Crop Protection group has a track record in modelling pest, including stripe rust and wheat bulb fly, and devising pest management solutions.
MT is an expert to provide cutting edge data analytic solution with machine learning techniques. Especially, MT have developed an interactive mobile-cloud toolkit for temporal-spatial data analytic toolkit.
Productions & Impact
- Open-source business model that provides basic mobile applications to small-size growers and public for free.
- The business-to-business approach, which sells professional software licences, services and professional training to middle-size growers and agronomists.
This project impacts on enhancing the UK’s leading position of global Agri-tech market. Winter wheat is one of the most important crops grown over the UK, however sustainable wheat protection is always under pression from fewer chemical inputs and reduced soil erosion. A timely and accurate inspection regime supported by the technology developed by this project will have a significant impact on this sector. The technology will be attractive to many actors in this area because of its practicability and usefulness. There will certainly be a market for this technology which will be a driving factor for more investment in future and creation of new commercial ventures.
