High Throughput Phenotyping for Energy Sorghum
We are developing a complimentary suite of robotic data-gathering tools that include autonomous aerial vehicles, ground-based robots, and distributed sensor networks to collect high-precision, high-resolution measurements of plants and environment in field settings throughout the crop life cycle. This information will be used to accelerate breeding programs for sorghum for use as an alternative energy crop. This approach aims for comprehensive daily coverage while also incorporating slower forms of data gathering that will provide unique phenotypic information critical to efficient breeding. An autonomous aerial vehicle will provide comprehensive downward-looking coverage, flying multiple missions per day that provide geometry and multi-spectral imagery. Small, ground-based robots will travel between the rows to take close-proximity measurements, generating high-resolution imagery and 3D models of individual plants. Additional ground robots will be equipped with a robotic arm that can be used to apply plant contact sensors such as a leaf porometer or a rind penetrometer. A distributed sensor network will collect environmental data continuously. Collected datasets will be combined with genetic data, and machine learning algorithms will be developed to better understand the relationships between plant genetics, early season phenotypes, and plant productivity.
kantor [AT] ri.cmu.edu
nuske [AT] cmu.edu