Proceedings of 2014 IEEE Information Reuse and Integration (IEEE IRI'14). Li Tan and Riley Wortman. IEEE Press. San Francisco, CA. August, 2014.
Yield mapping visualizes yield rate per geological distribution. It is frequently used as a baseline metric to measure yield efficiency in precision farming. A major challenge in mapping yield for specialty crops is how to collect accurate yield data without incurring substantial overhead to a farming operation. We design a yield efficiency analysis system that uses a cloud-based computing platform to acquire and analyze yield data. By reusing labor data collected by a cloud-based labor monitoring system that we developed earlier, our system calculates yield data from labor data, and computes yield map in real time and without the overhead for data acquisition. A distinctive feature of our approach is the introduction of a customizable yield distribution function that quantifies the probability of geographic distribution of fruits weighted at a Labor Monitoring Device. Practitioners may define yield distribution functions based on operational characteristics of an orchard, enabling our system adaptive for a variety of orchards with different harvesting operations and canopy architecture. Using a multi-tenancy software architecture, our system can support multiple orchards concurrently with improved scalability and data privacy. Our system has been deployed and tested on Amazon Web Services (AWS).