About

Girish Chowdhary is associated with the Center for Digital Agriculture at the University of Illinois. The center focuses on advancing digital and precision agriculture through research, education, and industry collaboration. The center's initiatives include developing AI-driven tools such as CropWizard, a decision-support service powered by generative AI designed for agricultural professionals, and supporting projects like AI AgriBench to evaluate AI tools in agronomy. The center also offers interdisciplinary educational programs, including a fully online Master’s Degree in Engineering with a concentration in Digital Agriculture, aimed at cultivating expertise in digital agriculture technologies. The Center for Digital Agriculture is engaged in research that spans data collection, storage, transmission, and analysis, with the goal of optimizing various aspects of agriculture such as precision farming, water use, and food manufacturing. Key projects include the development of datasets like PigLife for livestock industry optimization and exploring generative AI applications like CropWizard and CropGPT to enhance decision-making for farmers. The center actively promotes global perspectives on digital and smart agriculture through joint seminar series with institutions like National Taiwan University and hosts events such as the CDA Conference and AgTech Week to foster innovation and knowledge exchange in the field.

Research topics

• Computer Science
• Artificial Intelligence
• Engineering
• Machine Learning
• Computer vision
• Natural resource economics
• Business
• Ecology
• Geography
• Environmental science
• Economics
• Real-time computing
• Engineering management
• Mathematics
• Environmental resource management
• Systems engineering
• Agronomy

Selected publications

• Physics-informed neural network based damage identification for truss railroad bridges

  Structure and Infrastructure Engineering · 2026-02-13 · 1 citations

  preprintOpen access

  Railroad bridges are a crucial component of the U.S. freight rail system, which moves over 40 percent of the nation's freight and plays a critical role in the economy. However, aging bridge infrastructure and increasing train traffic pose significant safety hazards and risk service disruptions. The U.S. rail network includes over 100,000 railroad bridges, averaging one every 1.4 miles of track, with steel bridges comprising over 50% of the network's total bridge length. Early identification and assessment of damage in these bridges remain challenging tasks. This study proposes a physics-informed neural network (PINN) based approach for damage identification in steel truss railroad bridges. The proposed approach employs an unsupervised learning approach, eliminating the need for large datasets typically required by supervised methods. The approach utilizes train wheel load data and bridge response during train crossing events as inputs for damage identification. The PINN model explicitly incorporates the governing differential equations of the linear time-varying (LTV) bridge-train system. Herein, this model employs a recurrent neural network (RNN) based architecture incorporating a custom Runge-Kutta (RK) integrator cell, designed for gradient-based learning. The proposed approach updates the bridge finite element model while also quantifying damage severity and localizing the affected structural members. A case study on the Calumet Bridge in Chicago, Illinois, with simulated damage scenarios, is used to demonstrate the model's effectiveness in identifying damage while maintaining low false-positive rates. Furthermore, the damage identification pipeline is designed to seamlessly integrate prior knowledge from inspections and drone surveys, also enabling context-aware updating and assessment of bridge's condition.

  PublisherOA PDFDOI

• HyReach: Vision-Guided Hybrid Manipulator Reaching in Unseen Cluttered Environments

  arXiv (Cornell University) · 2026-03-22

  articleOpen accessSenior author

  As robotic systems increasingly operate in unstructured, cluttered, and previously unseen environments, there is a growing need for manipulators that combine compliance, adaptability, and precise control. This work presents a real-time hybrid rigid-soft continuum manipulator system designed for robust open-world object reaching in such challenging environments. The system integrates vision-based perception and 3D scene reconstruction with shape-aware motion planning to generate safe trajectories. A learning-based controller drives the hybrid arm to arbitrary target poses, leveraging the flexibility of the soft segment while maintaining the precision of the rigid segment. The system operates without environment-specific retraining, enabling direct generalization to new scenes. Extensive real-world experiments demonstrate consistent reaching performance with errors below 2 cm across diverse cluttered setups, highlighting the potential of hybrid manipulators for adaptive and reliable operation in unstructured environments.

  PublisherOA PDF

• Algorithmic design and implementation considerations of deep MPC

  ArXiv.org · 2025-11-21

  preprintOpen accessSenior author

  Deep Model Predictive Control (Deep MPC) is an evolving field that integrates model predictive control and deep learning. This manuscript is focused on a particular approach, which employs deep neural network in the loop with MPC. This class of approaches distributes control authority between a neural network and an MPC controller, in such a way that the neural network learns the model uncertainties while the MPC handles constraints. The approach is appealing because training data collected while the system is in operation can be used to fine-tune the neural network, and MPC prevents unsafe behavior during those learning transients. This manuscript explains implementation challenges of Deep MPC, algorithmic way to distribute control authority and argues that a poor choice in distributing control authority may lead to poor performance. A reason of poor performance is explained through a numerical experiment on a four-wheeled skid-steer dynamics.

  PublisherOA PDFDOI

• Non-destructive detection of pre-incubated chicken egg fertility using hyperspectral imaging and machine learning

  Smart Agricultural Technology · 2025-02-23 · 16 citations

  articleOpen access

  The egg industry heavily relies on accurate detection of egg fertility to optimize hatchery operations. Conventional methods, such as candling, rely on human interpretation, which is labor-intensive, time-consuming, and thus not efficient in large-scale operations. This study developed a fast, accurate, and non-destructive method of pre-incubated chicken egg fertility detection using hyperspectral imaging (HSI) and machine learning. The Extreme Gradient Boosting (XGBoost), Categorical Boosting (CatBoost), Random Forest (RF), and Support Vector Machine (SVM) calibration models were developed at full wavelengths (374–1015 nm), and the performance of the models was evaluated by 10-fold cross-validation and independent validation. Different spectral pre-processing and important feature selection methods were assessed for robust prediction model development. In addition to raw or non-synthetic data, synthetic data is also used to develop classification models. Using full wavelengths, the CatBoost model with synthetic data showed the best classification performance, attaining 95.1% accuracy in independent validation. The CatBoost models with fewer important features showed good prediction performance, making them computationally efficient, robust, and interpretable. The Shapley explainable artificial intelligence (AI) method was used to interpret the robust CatBoost model, revealing that wavelength regions associated with yolk color, pre-incubation cellular activities related to embryonic development, changes in hydration levels, and variations in protein and lipid contents between fertile and infertile eggs are crucial for pre-incubation chicken egg fertility classification. This study highlighted the efficacy of HSI combined with machine learning as a potential green technology for egg fertility detection towards a sustainable egg industry.

  PublisherDOI

• BYON: Bring Your Own Networks for Digital Agriculture Applications

  ArXiv.org · 2025-02-03

  preprintOpen access

  Digital agriculture technologies rely on sensors, drones, robots, and autonomous farm equipment to improve farm yields and incorporate sustainability practices. However, the adoption of such technologies is severely limited by the lack of broadband connectivity in rural areas. We argue that farming applications do not require permanent always-on connectivity. Instead, farming activity and digital agriculture applications follow seasonal rhythms of agriculture. Therefore, the need for connectivity is highly localized in time and space. We introduce BYON, a new connectivity model for high bandwidth agricultural applications that relies on emerging connectivity solutions like citizens broadband radio service (CBRS) and satellite networks. BYON creates an agile connectivity solution that can be moved along a farm to create spatio-temporal connectivity bubbles. BYON incorporates a new gateway design that reacts to the presence of crops and optimizes coverage in agricultural settings. We evaluate BYON in a production farm and demonstrate its benefits.

  PublisherOA PDFDOI

• Correction to: CROW: A Self-Supervised Crop Row Navigation Algorithm for Agricultural Fields

  Journal of Intelligent & Robotic Systems · 2025-11-13

  articleOpen access

  1. Figure 2: Replaced with the updated version containing the corrected dynamic function f_d. 2.

  PublisherOA PDFDOI

• Learning to Walk with Less: a Dyna-Style Approach to Quadrupedal Locomotion

  ArXiv.org · 2025-09-08

  preprintOpen access

  Traditional RL-based locomotion controllers often suffer from low data efficiency, requiring extensive interaction to achieve robust performance. We present a model-based reinforcement learning (MBRL) framework that improves sample efficiency for quadrupedal locomotion by appending synthetic data to the end of standard rollouts in PPO-based controllers, following the Dyna-Style paradigm. A predictive model, trained alongside the policy, generates short-horizon synthetic transitions that are gradually integrated using a scheduling strategy based on the policy update iterations. Through an ablation study, we identified a strong correlation between sample efficiency and rollout length, which guided the design of our experiments. We validated our approach in simulation on the Unitree Go1 robot and showed that replacing part of the simulated steps with synthetic ones not only mimics extended rollouts but also improves policy return and reduces variance. Finally, we demonstrate that this improvement transfers to the ability to track a wide range of locomotion commands using fewer simulated steps.

  PublisherOA PDFDOI

• Breaking the field phenotyping bottleneck in maize with autonomous robots

  Communications Biology · 2025-03-21 · 14 citations

  articleOpen accessSenior author

  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 (GxExM) on crop performance is time-consuming, expensive, and a major bottleneck to yield advancement. We demonstrate that an 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 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. 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. The robot measured these phenotypes with high accuracy and reliability, at scales sufficient to dissect interactions between genotypes and nitrogen rates in several environments. The results show that within-row, autonomous field robots hold great promise to increase GxExM understanding and decrease the amount of human labor required for plant phenotyping.

  PublisherOA PDFDOI

• Towards Efficient Large Scale Spatial-Temporal Time Series Forecasting via Improved Inverted Transformers

  ArXiv.org · 2025-03-13

  preprintOpen accessSenior author

  Time series forecasting at scale presents significant challenges for modern prediction systems, particularly when dealing with large sets of synchronized series, such as in a global payment network. In such systems, three key challenges must be overcome for accurate and scalable predictions: 1) emergence of new entities, 2) disappearance of existing entities, and 3) the large number of entities present in the data. The recently proposed Inverted Transformer (iTransformer) architecture has shown promising results by effectively handling variable entities. However, its practical application in large-scale settings is limited by quadratic time and space complexity ($O(N^2)$) with respect to the number of entities $N$. In this paper, we introduce EiFormer, an improved inverted transformer architecture that maintains the adaptive capabilities of iTransformer while reducing computational complexity to linear scale ($O(N)$). Our key innovation lies in restructuring the attention mechanism to eliminate redundant computations without sacrificing model expressiveness. Additionally, we incorporate a random projection mechanism that not only enhances efficiency but also improves prediction accuracy through better feature representation. Extensive experiments on the public LargeST benchmark dataset and a proprietary large-scale time series dataset demonstrate that EiFormer significantly outperforms existing methods in both computational efficiency and forecasting accuracy. Our approach enables practical deployment of transformer-based forecasting in industrial applications where handling time series at scale is essential.

  PublisherOA PDFDOI

• Precision Harvesting in Cluttered Environments: Integrating End Effector Design with Dual Camera Perception

  2025-05-19 · 3 citations

  article

  Due to labor shortages in specialty crop industries, a need for robotic automation to increase agricultural efficiency and productivity has arisen. Previous manipulation systems harvest well in uncluttered and structured environments. High tunnel environments are more compact and cluttered in nature, requiring a rethinking of the large form factor systems and grippers. We propose a novel co-designed framework incorporating a global detection camera and a local eye-in-hand camera that demonstrates precise localization of small fruits via closed-loop visual feedback and reliable error handling. Field experiments in high tunnels show that our system can reach 85.0% of cherry tomato fruit in 10.98s on average.

  PublisherDOI

Recent grants

• NRI: Collaborative Goal and Policy Learning from Human Operators of Construction Co¬-Robots

  NSF · $900k · 2015–2017

Frequent coauthors

• Jonathan P. How

  66 shared

• Eric N. Johnson

  Google (United States)

  38 shared

• Mateus V. Gasparino

  25 shared

• John Vian

  Boeing (Australia)

  21 shared

• John F. Quindlen

  Boeing (Australia)

  19 shared

• Hassan A. Kingravi

  Georgia Institute of Technology

  18 shared

• Nazım Kemal Üre

  18 shared

• Arun N. Sivakumar

  University of Illinois Urbana-Champaign

  17 shared