The participants in this program will experience various types of research and technology used to collect data for plant phenotyping in a field setting. The program will include traveling to field sites in Nebraska, working with researchers on projects to collect and analyze data, and regular interactions with the UNL team to discuss research projects and experiences.
The research mentor will work closely with each student to develop a plan to accomplish a specific research task for the project, such as sensor design, UAS image analysis, GIS analysis, and field data collection and analysis. Involvement in this program will help participants gain new experiences and strengthen their confidence in research. Our hope is that the students will obtain meaningful training and research experiences beyond repetitive field data collection or data entry to foster an interest in pursuing a career in agricultural research, extension, or production.
Competitive stipend: $6,000
Suite-style room and meal plan
Travel expenses to and from Lincoln
Campus parking and/or bus pass
Full access to the Campus Recreation Center and campus library system
Predicting grain and forage yield in small grain crops
The student will be involved in collecting manual and high-throughput phenotyping data in wheat and triticale yield trials. This will include collecting tiller counts, plant height, LAI and image/ hyperspectral sensor data to estimate grain yield, biomass quality, and forage yield in multiple locations in Nebraska. The student will work with graduate students, postdocs, faculty to analyze the data and develop models to predict forage and grain yield. These models will benefit the small grains breeding project by developing in-season tools for estimating important traits for variety release.
This project will allow the student to learn field research methods, participate in an active plant breeding program, develop new analysis and statistical skills as well as summarizing and presenting data to peers.
The student will work under one of the most advanced field phenotyping facilities in the country – NU-Spidercam. The student will have the opportunity to collect and work on a diversified set of high-throughput plant phenotyping data including images, spectroscopic reflectance, 3D point-clouds, and hyperspectral data cubes from corn and soybean.
The student will receive training to process these data using software like R or Python. The student will also be exposed to machine learning skills to model the plant traits (such as height, chlorophyll, LAI, stay green) from the high-throughput data.
Over the summer, a student will be involved in scoring one or more traits from both 1) a maize inbred diversity panel and 2) a replicated set of maize hybrid yield plots. Depending on the student's background and interests this may include manual measurements in the field and/or processing image and/or hyperspectral sensor data to estimate plant phenotypes.
The student will then use the trait data they themselves generated, combined with other traits, to build models to predict how the same maize lines performed in 2020 yield trials across 20+ locations in the USA using multiple machine learning models (random forest, support vector machine, extreme gradient boosting, etc.) and evaluate the relative importance of different plant traits to predicting yield in different environments. This project is expected to provide a student with an introduction to field research techniques, the Unix command line, and basic machine learning applications.
Drone-based Plant Phenotypes and Stress Sensing and Modeling
Over the summer, the student will work with an interdisciplinary team and be involved in generating, processing, and analyzing field-based plant phenotypic data for plant breeding and crop management purposes. Depending on the student’s background and interests, a project plan will be designed together by the student and the investigator and implemented by the student. This may include weekly drone-based flights to generate natural color, multispectral, and/or thermal imagery over the fields and ground-based plant trait measurements.
The student will then process the data/imagery collected to generate time-series field maps and extract key plant traits such as canopy height, ground cover, canopy temperature, vegetation indices, etc. These traits can be further modeled using classical plant growth models and/or statistical and machine learning models to assess plant stress levels or stress tolerance abilities, and/or derive management prescriptions.