Harris Nisar

Harris Nisar

PhD Candidate, Industrial Engineering · UIUC
Data Scientist · John Deere

I am a PhD Candidate in Industrial & Enterprise Systems Engineering at the University of Illinois Urbana-Champaign (UIUC) where I am advised by Professor Prof. Dušan M. Stipanović. My research focuses on developing artificial intelligence (computer vision/machine learning)and simulation-based (haptics/VR/AR) tools for patient interventions, human behavior modeling and training with applications in healthcare and agriculture. Broadly speaking, I am interested in how technologies can be used to improve outcomes in complex, high-stakes domains.

News

Feb 2026 📄 Our paper titled Translating Behavioral Interventions into Virtual Reality: The Transcend Framework for Immersive Health Design was accepted at Frontiers in Digital Health.
Jan 2026 📄 Our abstract entitled AI Models Facilitate the Automated Measurement of Social Gaze during Naturalistic Interactions has been accepted for an oral presentation at the International Society for Autism Research Annual Meeting (INSAR 2026). The link to the paper is coming soon!
Nov 2025 ✈️ I traveled to San Diego to present our paper DiffEye at NeurIPS 2025.
Oct 2025 🎉 I passed my PhD Prelimnary Exams! I'm happy to say that I'm officially a PhD Candidate!
Sept 2025 📄 Our paper DiffEye: Diffusion-Based Continuous Eye-Tracking Data Generation Conditioned on Natural Images was accepted at NeurIPS 2025.
Sept 2025 📄 Check out our preprint titled Robotic Trail Maker Platform for Rehabilitation in Neurological Conditions: Clinical Use Cases.
Jun 2025 🎉 I was awarded the ISE William A. Chittenden II Graduate Student Fellowship for AY 2025-26! This will fund my research for the final year of my PhD.
Jun 2025 🎉 I started working as part-time Data Scientist at John Deere. I'll be working on training neural networks to estimate field state like weed density from satellite images.
May 2025 🎉 I passed my PhD written exams!
May 2025 🎤 I gave a talk about how we implemented and tested a virtual reality simulation for umbilical venous catheter placement. See my slides here.
Mar 2025 📄 Our short paper titled Good Vibrations: Feeling and Interpreting Haptic Feedback was accepted to the International Conference of Learning Sciences (ICLS).
Mar 2025 🎉 I passed my PhD oral exams!
Feb 2025 📄 Our paper titled Assessing safety and feasibility of virtual reality intervention in patients with lung cancer: A pilot study was accepted at Supportive Care in Cancer.
Feb 2025 📄 Our brief communication titled Transfer validity testing of a virtual reality simulation for umbilical venous catheter placement was accepted at the Journal of Perinatology.
May 2024 ✈️ I traveled to Honolulu to present our paper Beyond the Screen: Gestural Perspective-Taking with a Biochemistry Simulation at CHI 2024. See our poster and video demo.
Feb 2024 📄 Our paper titled Beyond the Screen: Gestural Perspective-Taking with a Biochemistry Simulation was accepted at the CHI 2024 as a Late Breaking Work..
Feb 2024 📄 Our paper titled Development and usability of a virtual reality umbilical venous catheter placement simulator was accepted at the International Journal of Computer Assisted Radiology and Surgery.
Jan 2024 📄 Our paper titled Robotic Mirror Therapy for Stroke Rehabilitation through Virtual Activities of Daily Living  was accepted at the Computational and Structural Biotechnology Journal.

Experience

Pre-Doctoral Fellow
Industrial and Enterprise Systems Engineering, University of Illinois at Urbana-Champaign
January 2023 – Present
  • Recipient of ISE William A. Chittenden II Graduate Student Fellowship which fully covers my research appointment for one year (AY 2025-26).
  • Researching the learning dynamics of recurrent neural networks.
  • Developing a robotic platform for stroke rehabilitation in collaboration with occupational therapists at OSF St. Francis Hospital in Peoria, IL.
  • Using the platform for the collection of patient data and training machine learning models to differentiate between healthy and patient subjects and to forecast patient movement to provide assistance through the robot.
  • Taking technical courses in topics including machine learning, computer vision, and control theory.
Data Scientist (Part-Time)
John Deere, ISG, Machine Intelligence
June 2025 – April 2026
  • Leading the technical feasibility of predicting weed pressure from remote sensing data through deep learning to support John Deere’s precision weed sprayers (See & Spray).
  • Building PySpark pipelines combining times-series of satellite and equipment data from 100M+ acres, producing clean, leak-free training datasets (less than 2% duplicates, less than 5-hour runtime).
  • Enabling rapid training of deep learning models on large datasets (>1 billion rows) of remote sensing time-series and machine data using distributed training pipelines, Hydra configuration files and scripts to launch training runs via the Databricks Jobs API.
  • Presented model experimentation results to 150+ attendees at JDTechCon, John Deere’s premier internal technical conference.
  • Uncovered ~10% savings potential by fine-tuning See & Spray machine settings through the analysis of 100M+ acres of equipment data with PySpark and Pandas.
  • Determined optimal satellite provider for weed detection through analysis of 100M+ acres of equipment and imagery timeseries data (0.15 -> 0.33 correlation improvement over 40 days).
  • Created written and video documentation demonstrating how to use Databricks with VS Code, shared with 50+ data scientists and engineers.
Research Engineer
Healthcare Engineering Systems Center, UIUC
September 2024 – May 2025
  • Designed a new algorithm to generate realistic timeseries data (eye tracking) conditioned on natural images using diffusion and cross attention, resulting in DiffEye which was accepted to NeurIPS 2025.
  • Lead data annotation efforts with a team of 9 workers, using Python-based open-source tools to enhance machine learning models for behavioral data analysis on over 200,000 frames.
  • Developed data pipelines using Python (Pandas, NumPy, Matplotlib) to analyze video datasets, detect annotation outliers, and automate visualization pipelines, reducing processing time by 50%.
  • Implemented MLOps frameworks to easily launch and track experiments using tools like TensorBoard and Weights & Biases, increasing number of experiments launched from 2 a week to 1 a day.
  • Managed GPU computing resources of 2 high performing nodes with 8 GPUs in each for large-scale model training. Utilized these resources for distributed training, increasing experiment speed by 2x.
  • Facilitated IRB agreements to receive sensitive video of children performing clinical exams from 3 collaborators. Developing Data Use Agreements and pipelines to efficiently transfer the data (over 5 TB).
Simulation Engineer
Healthcare Engineering Systems Center, UIUC
September 2019 – September 2024
  • Developed 10 virtual and augmented reality simulations for education to enhance training effectiveness.
  • Led multi-disciplinary teams of engineers, clinicians, and artists, during full project lifecycle from ideation to deployment.
  • Conducted statistical analysis and applied time-series modeling techniques to enhance simulations, creating a data-driven feedback loop that improved user engagement metrics by over 50% during training sessions.
  • Managed over $400K in funding, producing 10+ publications and presentations on simulation-based learning.
Simulation Engineer
OSF St. Francis Hospital, Peoria, IL
September 2017 – September 2019
  • Developed a Unity-based mobile application for patient education (About Me 3D, formerly Rube-E), integrating Firebase for backend storage and a React.js-based content management system to enable real-time content updates.
  • Designed and built an internal web platform using HTML, CSS, and JavaScript to document simulation development workflows, standardizing system creation and delivery. Presented website to leadership receiving positive feedback.
  • Managed and mentored over 30 summer interns, overseeing project logistics, providing technical guidance, and fostering professional development.

Publications & Projects

Frontiers in Digital Health 2026
Frontiers in Digital Health
Translating Behavioral Interventions into Virtual Reality: The Transcend Framework for Immersive Health Design
Rosalba Hernandez, Killivalavan Solai, Soonhyung Kwon, Prasakthi Venkatesan, Drew Fast, Sandraluz Lara-Cinisomo, Harris Nisar

We propose a methodological framework that provides a practical roadmap for developing and optimizing VR-based behavioral health interventions.

INSAR 2026
INSAR 2026
AI Models Facilitate the Automated Measurement of Social Gaze During Naturalistic Interactions
Xu Cao, Frank Yang, Vipin Gunda, Fiona Ryan, Harris Nisar, Audrey Southerland, Elysha Clark-Whitney, Eliana L. Ajodan, Melaine R. Somekh, Elizabeth Stubbs, Chanel Miller, Catherine Lord, Rebecca M. Jones, So Hyun Kim, Agata Rozga, James M. Rehg

Our approach demonstrates that off-the-shelf vision foundation models can be simply converted into effective predictors of children's gaze targets during naturalistic social interactions. I led the data annotation efforts for this work.

DiffEye
NeurIPS 2025
DiffEye: Diffusion-Based Continuous Eye-Tracking Data Generation Conditioned on Natural Images
Ozgur Kara*, Harris Nisar*, James M. Rehg (* equal contribution, ordered by last name)

We propose DiffEye, a diffusion-based generative model for creating realistic, raw eye-tracking trajectories conditioned on natural images, which outperforms existing methods on scanpath generation tasks.

Robotic Trail Maker
Under Review
Robotic Trail Maker Platform for Rehabilitation in Neurological Conditions: Clinical Use Cases
Srikar Annamraju*, Harris Nisar*, Dayu Xia*, Shankar A Deka, Anne Horowitz, Nadica Miljković, Dušan M Stipanović (* equal contribution, ordered by last name)

We propose a robotic platform designed for rehabilitation in neurological conditions where patients complete trajectories with a robotic device. We used this platform to collect data from 10 stroke patients and 11 healthy subjects and found signficant differences between these populations when using our platform and trained deep learning models to classify these populations and to forecast their movements.

Jovialityr
Supportive Care in Cancer
Assessing safety and feasibility of virtual reality intervention in patients with lung cancer: A pilot study
Rosalba Hernandez, Harris Nisar, Thenkurussi "Kesh" Kesavadas, Mackenzie C McGee, Gregory J Gerstner, Angela Martinez, Carter Boyce, Sadia Anjum Ashrafi, Elizabeth L Addington, Alicia K Matthews, Safa Elkefi, Judith T Moskowitz

We developed a customized virtual reality (VR) software for people with lung cancer and evaluated its safety, acceptability, and preliminary efficacy, focusing on its potential to enhance mental health in patient care.

Development and usability of a virtual reality umbilical venous catheter placement simulator
International Journal of Computer Assisted Radiology and Surgery
Development and usability of a virtual reality umbilical venous catheter placement (UVC) simulator
Taylor Gohman, Harris Nisar, Avinash Gupta, M. Jawad Javed, Nicole Rau († corresponding author)

We developed a VR simulator for UVC placement, based on a participatory design approach to gather clinical requirements and iteratively incorporate clinical feedback. Experts (n=14) reported that the VR simulator provided a safe environment to make mistakes and the majority recommended this simulator to trainees. In a seperate study about the transfer validity of our simulator, we also found comparable training levels when comparing our VR simulator to traditional video based methods.

Beyond the Screen: Gestural Perspective-Taking with a Biochemistry Simulation
Extended Abstracts of the CHI Conference on Human Factors in Computing Systems
Beyond the Screen: Gestural Perspective-Taking with a Biochemistry Simulation
Taehyun Kim, Harris Nisar, Robb Lindgren, Jiahao Zhang, Xiaoyu Tang, Matthew Lira, Aishwari Talhan

We developed a biochemistry simulation to teach about electrochemical membrane potential where the learner uses gestural interactions while wearing a haptic glove to explore and understand complex biochemical processes. We conductged a user study and found that students interpreted the scientifc concepts effectively and that the gesture interactions enhanced their interpretation of the simulation.

Robotic mirror therapy for stroke rehabilitation through virtual activities of daily living
Computational and Structural Biotechnology Journal
Robotic Mirror Therapy for Stroke Rehabilitation through Virtual Activities of Daily Living
Harris Nisar, Srikar Annamraju b, Shankar A. Deka c, Anne Horowitz d, Dušan M. Stipanović b

We developed a 3D robotic mirror therapy system for upper limb rehabilitation of hemiplegic stroke patients through virtual activities of daily living. The patient controls one robot with their unaffected limb, while the other robot mirrors the movements to provide therapeutic feedback to the affected limb. The unaffected limb is guided by the therapist through a third robot.

Education

PhD, Industrial & Enterprise Systems Engineering
University of Illinois Urbana-Champaign
January 2022 - July 2026 (expected)
MS, Industrial Engineering (Advanced Analytics)
University of Illinois Urbana-Champaign
January 2020 - December 2022
BS, Bioengineering
University of Illinois Urbana-Champaign
August 2013 - May 2017