Identifying and Controlling Weeds - OSCAR, India
The Open Source Simple Computer for Agriculture in Rural Areas (OSCAR) project involves the prototyping of an application software for weed identification and control of the rice and wheat crop systems of the Indo-Gangetic Plains. It is targeted at being deployed on low-cost computing devices running GNU/Linux that can be shared among farmers of a local community.
Summary
The Open Source Simple Computer for Agriculture in Rural Areas (OSCAR) project from the French Institute of Pondicherry (IFP), involves the prototyping of an application software for weed identification and control of the rice and wheat crop systems of the Indo-Gangetic Plains (IGP). It is targeted at being deployed on low-cost computing devices running GNU/Linux that can be shared among farmers of a local community.
OSCAR has been developed for both desktop computers and Simputers using Free and Open Source Software (FOSS) tools, and is released as FOSS. FOSS is chosen as it is important to make the software freely available to the farmers and to encourage contributions to OSCAR from various groups and organizations in the IGP. These groups help enhance the system by cataloguing the weed species in their region and their respective control measures. It is also important to be able to perform customizations to suit local languages and cultural practices as the content of the application, particularly, the control measures of weeds may vary with the practices in different regions.
OSCAR is unique in that it is the first of its kind within the domain of information and communications technology (ICT) applications for agriculture. By being available as FOSS, it promotes the aggregation of information from academic/research institutions as well as from traditional knowledge systems.
The project has tested the application with various target groups in the four countries of the IGP – Bangladesh, India, Nepal and Pakistan, with encouraging results. A key lesson learned from the project is that any implementation of ICT applications for a rural agrarian population has to take into account the prevalent basic issues in agricultural practices and infrastructural constraints. It is essential that methods to integrate participation from the local communities are considered.
The current project is concerned with delivering and testing the prototype OSCAR application. While this has been largely successful, the ability to move from this to the actual working environment at the field level remains uncertain, as is the ability to continue to consolidate and sustain the enthusiasm and efforts of the different agencies in the different countries to participate, contribute, build and enhance on the existing set-up.
The actual deployment of OSCAR will probably not be done on Simputers as several issues with implementing and using it on a Simputer were experienced during the course of the development of OSCAR and its subsequent trial runs. To overcome this and to enable it to reach a wider audience, OSCAR has been successfully converted into a web-based application.
Background of Organization
The French Institute of Pondicherry, a multidisciplinary research centre located in Pondicherry, India carries out research and training activities in South and South-East Asia in the fields of Indology, Social Sciences and Ecology. The IFP is highly regarded for its work in the study of Sanskrit and Tamil languages, literature and grammatical traditions (in collaboration with the Ecole Française d’ Extrême Orient) as well as for its vegetation mapping of South Indian Forests and bioclimates in South-East Asia.
The Laboratory of Geomatics and Applied Informatics in the Institute specializes in Geographical Information Systems (GIS) and in ICT. It is responsible for all the ICT infrastructural needs of IFP as well as the development of databases and multimedia products for the departments at the Institute. The Laboratory is also in charge of original applications development, aimed at improving access to new technologies.
Objectives of Project
The OSCAR project aims to address the issue of declining agricultural productivity in South Asia by producing a tool for decision-making in weed identification and control. The specific objective of the project is to demonstrate a prototype of this tool implemented in software and running on desktop computers and low-cost computing devices.
This decision-making tool targets farmers of the IGP at the village level and it will be available in different local languages so as to address the cultural diversity of the project area. An existing software for species identification - IDentification Assiste pour Ordinnateur (IDAO) is ported to a FOSS platform and deployed on low-cost computing devices running GNU/Linux, that can be shared among farmers of a local community. A model database for identification and control of a set of the most important weeds of the IGP is developed for integration with the species identification software. The usage, appropriateness and acceptance of the tool at the farmer community level are also assessed.
The primary target groups are the farmers and village communities of selected villages from the four countries of the IGP – Bangladesh, India, Nepal and Pakistan. The indirect target group is the farming community of the region in general. The ultimate end users of the OSCAR application will be the farmers, village communities and also all the actors involved in the support of agricultural activities, such as development workers, NGOs, training and extension officers of semi-government and government agencies, and students of botany and agronomy at the colleges and universities.
For this project, IFP is working with three partners from Asia and Europe:
- The Centre de Coopération Internationale en Recherche Agronomique pour le Développement of France specializing in development-oriented agricultural research for hot regions.
- The Rice and Wheat Consortium of Bangladesh, India, Nepal and Pakistan has weed science experts and a strong background in agriculture in the IGP.
- The Centre of Innovation Studies, Wageningen University from the Netherlands, is concerned with the way communication can be strategically brought in to reinforce development, pro-social behaviour, organizational efficiency, and collective decision-making. Its interests cover the broad and interconnected areas of agriculture, food and health, environment and nature, in western as well as in non-western countries.
FOSS Application
Description
The prototype application, OSCAR, is the primary outcome of the project. With a weed identification system at its core, it has information on 50 of the most common weed species for the rice and wheat crop systems of the IGP. OSCAR helps the farmer to properly identify a weed species and provides information on the botanical aspects and appropriate control measures.
Choice of FOSS
FOSS is chosen as it allows easy customization of the software and the data formats used are known and open. These considerations are important so as to encourage contributions to OSCAR from agronomists, researchers, student community and concerned development organizations in the IGP. These organizations can help to enhance the system by cataloguing the weed species in their region and their respective control measures. If the application is based on proprietary formats and software then the whole purpose of OSCAR will be defeated since for any future enhancements and customization, the current agencies involved will be required to implement them. Also, taking into consideration the cultural diversity of the target region that covers four countries with different languages and cultural practices, it will not be practical to have a common application for the whole region. Though efforts are being made to port OSCAR into respective local languages, the content of the application, particularly the control measures of weeds vary with context. So, by using FOSS, the community can customize the application to suit its own local needs. For example, the number of weed species for lowland rice crop systems in Bangladesh can be increased in the database compared to the numbers of weeds for wheat, so that the application contains more tangible and relevant information.
Development and Implementation
OSCAR is designed to work on GNU/Linux platforms. The Gtk graphics library package is used for developing the Graphical User Interface (GUI) for the application while the libgda library (a library for writing database programs) is used for database access. MySQL is utilized as the database backend with the gda-mysql package being deployed to interface between libgda and MySQL. OSCAR is being developed for both desktop computers and Simputers.
The core component in the OSCAR application is the species identification kit (identikit). The identikit contains a model plant species with different user selectable characters such as root, habit, leaves and flowers. Each character has a number of character states which are in turn selected by the user. Based on the user's selection, the identikit displays the most appropriate match compared to the species available on the database. The base code is written in the C programming language and the characters and character states are stored in the database. The database is independent of the base code so as to facilitate future enhancements to the database.
Deployment
The application is targeted towards the farmers, extension workers and students of the IGP. The project focuses mainly on making the software available on FOSS to the target groups. The application has been widely tested with farmers, extension workers, students and researchers and has received enthusiastic responses. The application is made freely available on the website and interested parties are encouraged to download and use the application. Any agency such as a government or NGO involved in rural development can make use of the application.
The project has tested the application with the various target groups in the four countries of Bangladesh, India, Nepal and Pakistan. It was found that the local communities were very open and ready to adopt and own the application. From the researchers and teaching community at the universities and research institutions to the development workers at grass-roots level, everyone was readily convinced about the advantages of the FOSS model and was very eager to actively contribute in one way or another.
The deployment of OSCAR is unique in that it is the first of its kind within the domain of ICT applications for agriculture. The nature of the application addresses one of the pressing needs of the target groups, namely, to identify correctly the weed species and control them effectively. By virtue of OSCAR being released as a FOSS application, participating parties can freely contribute towards building up its knowledge base. The aggregation of information is provided not only from academic/research institutions but also from traditional knowledge systems, for example, cultural practices in weed control like hand weeding.
Impact
Enthusiastic responses to OSCAR have been received, from the research community in particular, in all the target countries. The fact that they can customize the application to fit their own environment has invoked much interest in OSCAR.
OSCAR will encourage and result in the participation of local bodies and organizations, like panchayats (councils at the village level), cooperatives, farmers' associations, women's self-help groups, etc. to build their capacities with resources accessible within their village. This will also enable these micro-organizations to manage their resources at village level with the combined knowledge base at local and institutional levels. This effect can contribute enormously towards improving the decision-making capacity of the farmers, not only on issues related to farming, but also on other issues related to development as a spill-over effect.
Another impact is the enablement of partner organizations to garner expertise on various issues including different local needs, the approach and appropriateness of ICT aids in farming, technical presentation of the application, and the integration of technical information in local farming practices and traditional knowledge systems. These factors contribute to the design of better management techniques for the future.
Lessons Learned
The project has contributed much to the understanding of how to design and implement an ICT application for agriculture. One key lesson learned during the implementation of the project was that any ICT application targeted towards a rural agrarian population has to take into consideration the prevalent basic issues in agricultural practices and infrastructural constraints. Acceptance by the target group and methods to integrate participation from local communities is essential for any ICT intervention at the grass-roots level to be successful. Merely dishing out information without considering social issues and differences at the village level and existing channels for information will only make the initiative redundant.
Current Status of Project
The project has completed its final stage and final releases of OSCAR for GNU/Linux and the Simputer. A set of recommendations for ICT initiatives for agriculture has also been published.
During the course of the development of OSCAR and subsequent trial runs, the project found several issues with implementing and using OSCAR on a Simputer. These include the lack of a standard development platform, limited availability of storage space, high local pricing (prices are comparable to a desktop PC) and the poor hardware support locally. Given all these restrictions, it is unlikely that further development and deployment of OSCAR on the Simputer will take place. Instead it has been decided to port OSCAR to a web-based FOSS application as this will make it more visible and will enable it to reach a much wider population and area. This has been done successfully.
Benefits and Challenges
OSCAR provides farmers in the IGP with the information and ability to identify and control weeds in their paddy and wheat fields. Prior to this, it was very difficult for non-botanists to identify weeds properly and correctly. The OSCAR application, based on a GUI and uses drawings instead of technical jargon, is simple and easy to use.
The use of FOSS tools and applications to develop and build OSCAR as well as a FOSS database to store the data provides an open and adept environment to develop and standardize applications such as OSCAR. The scientific products of academic and research institutions can be deployed on FOSS thereby encouraging wider participation and dissemination.
Positive and spirited responses experienced during the testing of OSCAR shows that there is a real need for such applications and FOSS is the only way such initiatives can be sustained and replicated for a wider region. If OSCAR is not a FOSS application it may have been yet another centrally-managed decision-support system that would cease to exist once the project ends. There is now active participation from different agencies to contribute further towards enhancement and deployment.
However, in spite of the encouraging signs and activities witnessed as described above on participation in the OSCAR project, the biggest challenge still remains the ability to consolidate and sustain the enthusiasm and efforts of the different agencies in different countries to participate, contribute, build and enhance on the existing setup of the project. As the current project owners and drivers are research organizations and working within the framework of a project with a very specific objective of demonstrating a prototype, it can be difficult to fulfil the expectations raised at the field level.
Conclusion
The OSCAR prototype of an application software for weed identification and control of the rice and wheat crop systems of the IGP has been successfully developed as FOSS and tested in various target groups in the four countries of Bangladesh, India, Nepal and Pakistan. The usage of FOSS to develop and implement OSCAR makes it possible to encourage contributions to it from agronomists, researchers, student community and concerned development organizations in the region. With the deployment of OSCAR, farmers in the IGP will be able to better identify and control weeds in their paddy and wheat fields. An indirect impact of the project is the encouragement of local bodies and organizations to help themselves by building their capacities with resources from within their village leading to better resource management and contributing towards the decision-making capacity of the farmers on farming and other issues related to development.
Contact Information
Project: Open Source Simple Computer for Agriculture in Rural Areas (OSCAR)
Organization: French Institute of Pondicherry (IFP)
Contact Person: Balasubramanian D
Street / Postal Address: 11, St. Louis Street, Pondicherry, India 605001
Email: balu.d@ifpindia.org
Phone: +91 413 222 5820
Fax: +91 413 233 9534
Website
The OSCAR project website
http://www.oscarasia.org