For pest management, we propose an agricultural systems model employing environmentally friendly agricultural technology based on a Bayesian network. In this model, the network nodes represent prior/posterior pest, natural enemy, pesticide dispersion, environmental load, crop, and sustainability. The conditional probability table (CPT) of each node reflects partial order relations that, in turn, reflect agricultural technologies (pesticide dispersion). The simulation is repeated with updated probability values for the existence of prior/posterior pests through the Bayesian network. Eventually the probability values in each node converge to some fixed values. Model simulations allowed us to evaluate new agricultural systems and should eventually aid in the establishment of environmentally friendly agricultural technology systems and support decision-making in agriculture. The model's fit is achieved by examining partial order relations among phenomena and can be improved by obtaining new information or actual data. Although the CPT is made from subjective probability values, the repetition of improvements in the model is expected to provide practical decision-making support tools.

View full abstract

Sensor data are being used for environmental measurement at many research institutes. To exchange sensor data between measurement sites, it is necessary to open sensor data to the Internet with a Web server. However, measurement sites cannot always operate via a Web server because of operational and security issues. Furthermore, support for NAT (Network Address Translation) and many different types of sensors are also needed. For these reasons, we developed a system based on hybrid P2P technology to facilitate exchange of sensor data. We conducted experiments to evaluate the fundamental performance of the system and found that we were able to efficiently search for and obtain sensor data using the system. We were also able to share sensor data via the Internet in an NAT environment without the need to operate a Web server by installing a developed Node function. Furthermore, we implemented a function that allowed the user to easily add different types of sensors.

View full abstract

Both economic and environmental indicators are important in environmentally friendly farm planning systems. In such systems, the user will need to be able to store relevant information and share it with others. Our goal was to determine which information system would be suitable for decision-making for farm activities from both the economic and environmental viewpoints. Specifically, we developed a farm-planning decision support system that can simultaneously consider the environmental impact of agricultural chemical use and other factors. With this system, a farmer can (1) calculate the environmental risk indicator of agricultural chemical use, (2) estimate greenhouse gas emissions from agricultural production, and (3) manage his or her own database and make a farming plan with the help of the database. In addition, we enhanced the function of the Microsoft Excel workbook, which is used to input the information about the farming technology system. These system functions were developed and integrated into the FAPS-DB system. We confirmed that a user could simulate the impacts of a change in economic conditions on the economic and environmental performance of a farming plan by integrating our system with the FAPS-DB. The system simultaneously provides various quantitative and graphic management indicators for the user via the Internet. We also discuss how the simulation analysis produced by our system can support decision-making while considering the roles of members in an organization. We found that the system could be used as a farm planning tool to support environmentally friendly agriculture.

View full abstract

This paper describes the role of educational programs, institutional settings, and information and communication technology (ICT) tools on the development of integrated production in Almeria, Spain. To consider the necessity of integrated production and the role of above factors, the transfer of integrated production system to other countries is also discussed. Integrated Production (IP) is an agricultural production system that uses natural methods and natural mechanisms of production to manage pests and diseases, taking into account the protection of the environment and the farm economy, with social responsibility. IP not only takes into consideration operations at the farm level but also at packaging, processing, and labeling. IP regulations not only are in line with other certified quality norms such as Naturane, AENOR, and Global-GAP but also include more demanding regulations and provide users tools to assure product quality and production requirements. The availability of ICT tools such as administrative, technical support, and private management and traceability have significantly contributed to control and secure the accrue application of IP rules by the regional government and certification entities. They also facilitate the adoption of IP by farmers through the groups of integrated production (APIs), and to translate huge amounts of information in a comprehensive manner so that consumers are able to verify product quality and production requirements. IP has not only been provided of the normative framework but also of educational programs and institutional settings. Farmers are trained on pest, pesticides, and labor risk management while technicians are capacitated on a wider range including product quality certification norms at both farm and processing center levels. The development of ICT tools, the provision of educational programs along with the cooperation and coordinated work of all the stakeholders and organizations involved in IP is bringing about the successful application and development of this agricultural system.

View full abstract

Information such as tree structure, canopy extension, and leaf color are quite important for cultivation and management of trees in fruit orchards. Aerial images of the trees combined with field geophysical information are useful tools in this regard, so we developed a close range remote sensing system consisting of a wireless video camera and a 3-axis accelerometer. The system was thrown into the air by hand, and images were captured in various directions while the apparatus freely fell to ground. Only the images taken correctly facing the target area were automatically extracted from the video images with a support vector machine using 3-axis acceleration data. The results of the image selection analysis showed that the precision rate was 53.7% and the recall rate was 57.4%. Furthermore, system operation did not require sophisticated control devices or other manipulations.

View full abstract