As agriculture struggles to support the rapidly growing global population, the productivity and quality of food crops is being undermined by plant diseases and nutritional disorders. These oftentimes result in catastrophic losses, estimated to be between 20% and 40% for major food and cash crops (rice, wheat, maize, barley, potatoes, soybeans, cotton, and coffee) at country and regional levels in different continents (Oerke, 2006). Crop losses have grave implications for the wellbeing of rural families, as well as for food security and economic and political stability at the national, regional and global level (Zadoks and Schein, 1979; Savary and Willocquet, 2014; Avelino et al., 2015). To avoid losses, farmers spend billions of dollars on plant disease management and chemical pesticides and fertilizers, often without adequate technical support, resulting in poor control, pollution and other harmful effects. Plant diseases can also devastate natural ecosystems, further exacerbating environmental problems caused by habitat loss and poor land management.
Although agricultural extension workers and other professionals are responsible for the recognition of plant diseases, intelligent systems can facilitate in-situ diagnosis at early stages. Various systems and applications for specific crops, agricultural systems and even countries have been developed for this purpose. They rely on facts described by the user or on image processing of plant photos in visible/infrared light. The recognition of a disease or any many other stress factors (e.g. nutrient deficiency) is based on symptoms such as lesions or spots, which can be characterized based on color and shape (RGB images: true color image stored digitally as an m-by-n-by-3 data array that defines red, green, and blue color components for each individual pixel), spectral signatures and even temperature. Low cost and user-friendly smartphone applications with such capabilities may help farmers to early detect and even identify diseases and disorders in their crops in-situ. Higher cost molecular analyses and tests could follow if necessary.
For example, Plantix (https://plantix.net/en), developed by Hannover University-based startup Peat, is a free mobile application which offers farmers and gardeners the possibility to receive decision support directly on their smartphone. Using image recognition, the app is able to identify the plant type, as well as the appearance of a possible disease, pest or nutrient deficiency. Plantix takes advantage of deep learning technology which involves neural networks. Furthermore, it provides information on treatment and preventive measures. The daily new images sent by Plantix users worldwide to the system allow it to constantly learn and send back to the users up-to-date information and alerts in terms of plant diseases, pests and their worldwide distribution in real time. Other example is Maize Doctor (http://ww12.maizedoctor.org/), developed by CIMMYT (The International Maize and Wheat Improvement Center), which a simple, stepwise method for identifying maize production problems, pests and diseases and suggests ways you can overcome your problems. However, the weakness of these apps (and many other of similar characteristics) is that they depend to a large extent on a central server and lacks autonomy from the own capacities of the mobile phone.
The aforementioned weakness can be overcome by the system developed by the University of Barcelona (UB): to be exact, the Integrative Crop Ecophysiology Group within the Plant Physiology Unit, Department of Evolutionary Biology, Ecology and Environmental Sciences. It comprises a simple yet fully integrated data collection and processing platform that connects the Open Data Kit (ODK) Collect (https://opendatakit.org/) with a Google cloud-based organization structure for gathering image and questionnaire data that can be easily accessed and processed in real-time using any personal computer connected to the internet (for more details see https://integrativecropecophysiology.com/fusion/). The ODK Fusion application allows for even easier data-sharing via GoogleSheets and geographical visualization and machine learning analyses through the new Google BigQuery GIS tools. Both may be used along with the mobile app geographical data as inputs for driving satellite-based analyses through the Google Earth Engine API. Together, they not only enable strong user feedback and an enlightening user experience that may contribute to better participation, but also a very strong potential for backend analyses at local to regional scales.
The International Center for Biosaline Agriculture (ICBA), based in Dubai, proposes to implement a project in partnership with UB that will take advantage of the system developed by the latter to create a low-cost, user-friendly application for smartphones that farmers can use to identify and address diseases and nutritional disorders in their produce and thus minimize yield losses. ICBA has previously developed several brochures and booklets to help farmers detect some of the problems affecting their crops. However, as the global population becomes more accustomed to and reliant on rapidly developing ICT technologies, farmers – particularly new generations – are more attracted to digital applications than paper-based publications. The almost universal availability of mobile phones, even among farmers in developing countries, provides many with the opportunity to access the information anywhere.
The project will initially target farmers in four countries of the Middle East and North Africa (MENA) region, including the United Arab Emirates (UAE), Egypt, Tunisia and Morocco, and will be expanded to other countries in the future. The application will first focus on the greenhouse crops commonly grown in the region, such as tomato, capsicum and cucumber, as well as some newly introduced crops. Among them is quinoa, a stress-tolerant crop with the potential to make a significant contribution to food security in the countries of the MENA region, particularly those adversely affected by climate change impacts such as water stress and increasing soil and water salinity. The application will be gradually expanded to include a wide range of other conventional and underutilized crops.