Modular System Architecture for Individual Plant Phentotyping with an Autonomous Field Robot

Abstract

A concept of a modular system architecture for plant phenotyping with an autonomous robot has been developed and realized. The system takes into account the inclusion of several sensors - placed at different positions - with a broad range of data volume and data rates. The database confirms the online data storage of sensor data including position and times steps with RTK-DGPS accuracy, thereby allowing single plant analysis. Moreover the configuration of the gateways and sensors and the error handling are other tasks of the database. The human-machine interface allows a flexible setting of parameters and the inclusion of additional sensors. The data of the spectral and morphological sensor and positioning information can be accessed by standard software tools for plant phenotyping analysis. Moreover, as part of the architecture for an autonomous field robot the system has a high potential for further applications in this field.