Breaking the field phenotyping bottleneck in maize with autonomous robots
Data files
Mar 11, 2025 version files 414.88 KB
-
Data_for_Case_studies_1_to_5.xlsx
389.98 KB
-
Data_for_Figure_S2_Trait_Validation.xlsx
22.63 KB
-
README.md
2.27 KB
Abstract
Understanding phenotypic plasticity in maize (Zea mays L.) is a current grand challenge for continued crop improvement. Measuring the interactive effects of genetics, environmental factors, and management practices such as nitrogen rate (GxExM) on crop performance is time-consuming, expensive, and a major bottleneck to continued yield advancement. We demonstrate that a novel autonomous robotic platform, capable of collecting biologically relevant and commonly measured phenotypes, within a maize canopy at high-throughput, low-cost, and high-volume is now a reality. Field teams used multiple TerraSentia autonomous ground robots developed by EarthSense, Inc. (Champaign, IL) to capture data using a suite of low-cost sensors from nearly 200,000 experimental units, located at 142 unique research fields in the USA and Canada, across five years. Novel computer vision and machine learning algorithms, developed by EarthSense, Inc., analyzed these in-canopy multi-sensor data to deliver ground-truth validated plant height, ear height, stem diameter, and leaf area index at multiple time points during each season. We show the robot measured these phenotypes with high accuracy and reliability, at scales sufficient to functionally dissect interactions between genotypes and nitrogen rates in several environments. The results show that within-row, autonomous field robots hold great promise to increase understanding of GxExM interactions in maize research and decrease the amount of human labor required for plant phenotyping.
https://doi.org/10.5061/dryad.m905qfvb6
Description of the data and file structure
These files contain data collected either by the TerraSentia robot or by manual measurements for validation. Data is organized by the Figures presented in paper. Each dataset includes metadata on year, location and date of collection, field coordinates (X-y grid) for each plot, the maize hybrid planted in each plot, and the phenotypic data values.
Files and variables
File: Data_for_Figure_S2_Trait_Validation.xlsx
Description: Data output from TerraSentia rover and ground-truth measures from the same plots, used to demonstrate accuracy of rover-estimated phenotypes.
Variables
- Leaf Area Index (LAI), Stem Diameter, Plant Height, Ear Height, timestamp, location of plot in field column (x) and range (y) grid.
File: Data_for_Case_studies_1_to_5.xlsx
Description: Data for each figure in the article.
Data corresponding to Table 1 and Figures 2, 3, 4 and 5 are provided in a separate spreadsheet tab.
Table 1 = Observations for Ear Height from a population of pre-commercial hybrids using the TerraSentia robot at five locations in 2022.
Figure 2 = Plant canopy traits collected by the TerraSentia robot from an experiment investigating hybrid by plant density by level of nitrogen fertilization.
Figure 3 = Measures of nitrogen utilization and its component traits among maize hybrids grown at the Illinois N-responsive field site.
Figure 4 = Data collection from an on-farm experiment assessing variation in hybrids supplied different rates of nitrogen fertilizer, at two experimental locations.
Figure 5: Robot data collection from precommercial breeding trials of different hybrids grown at multiple locations in 2022.
Variables
- Descriptions of data variables are provided in each tab. Each data table contains information about maize hybrids, location of trial, location of plot within field as x-y grid of range-plot, field management treatments (N fertilizer rate, plant density), and phenotypic values.
- Blank cells under individual traits represents missing data
The TerraSentia rover collected RGB video images and LIDAR point clouds from corn plants as it moved between rows in the field. Video data was processed by algorithms described in the paper to estimate either ear height, stem width, or leaf area index. Ground-truth phenotypes were collected from a subset of the same plots using established methods: plot combine to estimate grain yield, plant and ear height using a centimeter-scaled measuring pole, stem width with calipers, and leaf area index by manual measurement and a LICOR LI-3100C scanner. Both rover and manual measurements were from multiple plants per plot, and reported as means of individual measures per plot.