Adaptive Data Collection for Rapid Evaluation of New Plant Varieties

World population is projected to reach 9.6 billion by 2050, and yields of most staple crops are not increasing at a fast enough rate to meet the corresponding nutritional needs. The overall goal of this research is to create new CPS science and technology for adaptive data acquisition for high throughput plant phenotyping, which will accelerate breeding progress for high yield food crops. Currently, the primary bottleneck in identifying high-yielding crops is the limited number of plants a breeder can evaluate in a growing season. We develop algorithms that can be used by robotic technology to very rapidly collect data in breeding sites so as to guide breeders in decisions about selection of high performance crop varieties. We focus on sorghum breeding as an example, but the proposed techniques will be generalizable to other agronomic crops as well as many other environmental monitoring tasks that can benefit from integrating temporal and spatial sensing and physical sampling. The research leverages the expertise of plant geneticists and breeders and will be evaluated in simulation and in a proof-of concept demonstration with real robotic platforms.