Since citrus is often grown on terraced orchards in west Japan, their qualities and yields are various, because of different environments such as temperature, humidity, solar radiation, precipitation, soil condition, tree age, and other conditions on different elevation fields. In this study, the concept of a mobile fruit grading robot and measurements from the robot are described. This robot consists mainly of a roller pin conveyor, two machine vision systems, an NIR inspection system, a GPS, and a travelling device. One or two operators can operate this robot for harvesting, because the robot keeps a constant distance between operators and the robot by laser sensors as it travels. Harvested fruit are put into the roller pin conveyor through two gates, as information of growth in northern or southern canopy is added for the fruit canopy location. One vision machine inspects harvested fruit, while the other monitors tree leaf color and canopy size. An NIR inspection system measures internal sugar content and a GPS records robot location. It is possible to accumulate fruit color, size, shape, defect, sugar content, tree leaf color, and canopy size into a database corresponding to tree location. Adding to this robot's information, analysis of satellite images is conducted and field-server information is collected. The producers' operation records can be input to the database with the satellite and field-server information. This database is used both for farming guidance to conduct tree management and for food traceability in relationship to safety and security in food distribution.