Writing
Papers and essays
Selected work, newest first. Each entry carries one plain-language sentence about what it actually claims. The complete list is on Google Scholar.
2026
-
Kernel dynamics applied to terrain represented as graphs: how the distinctions a planetary-surface model can make evolve, and which topological quantities are conserved along the way.
-
The spectral form of the kernel-dynamics program — where trajectories through kernel space settle, how they move between fixed points, and when they change regime abruptly.
-
Treats the kernel — the structure determining what an agent can distinguish — as a dynamical variable governed by Maximum Caliber, with a Landauer-type lower bound on the work that conceptual change costs.
-
How Navagunjara Reborn was designed in Arizona, crafted in Odisha, and assembled once on the playa — and why the final form is best read as a maximum-entropy configuration satisfying every contributor's constraints at once.
-
Essays on building, loss, documentation, and entropy — written while the sculpture was paperwork.
2022 – 2024
-
Virtual Shake Robot: Simulating Dynamics of Precariously Balanced Rocks for Overturning and Large-displacement Processes
A simulation framework for how fragile rock formations topple, giving constraints on the ground motion a site has — and hasn't — experienced.
-
Shakebot: A Low-cost, Open-source Robotic Shake Table for Earthquake Research and Education
A shake table anyone can build, for classrooms and labs that can't afford the conventional kind.
-
A Survey of Decision-Theoretic Approaches for Robotic Environmental Monitoring
A map of how robots decide what to measure in the environment, organized by the decision theory underneath each method.
-
Quantifying and Analysing Rock Trait Distributions of Rocky Fault Scarps Using Deep Learning
Deep learning on drone imagery to measure the rocks on fault scarps, as a record of how the scarps formed.
2015 – 2021
-
OpenUAV Cloud Testbed: A Collaborative Design Studio for Field Robotics
Field robotics in the browser — a shared simulation and development environment for students and researchers.
-
Terrain-Relative Diver Following with Autonomous Underwater Vehicle for Coral Reef Mapping
An AUV that follows a diver using the reef itself as reference, so the mapping goes where the expert goes.
-
Robust Fruit Counting: Combining Deep Learning, Tracking, and Structure from Motion
Count every fruit in an orchard from video without counting any of them twice.
-
Counting Apples and Oranges with Deep Learning: A Data-Driven Approach
A learned pipeline for yield estimation from images, and one of the most-cited results in agricultural robotics.
-
Towards Autonomous Phytopathology: Outcomes and Challenges of Citrus Greening Disease Detection Through Close-Range Remote Sensing
Can a robot see a sick tree before a human can? Partly — and here is exactly where it fails.
-
Devices, Systems, and Methods for Automated Monitoring Enabling Precision Agriculture
The sensing platform behind the Penn agricultural robotics program, and the basis of a patent.
-
Data-Driven Robotic Sampling for Marine Ecosystem Monitoring
Let the AUV choose where to sample based on what it has already learned about the bloom, and the samples get better.
2009 – 2014
-
Hierarchical Probabilistic Regression for AUV-Based Adaptive Sampling of Marine Phenomena
A model that learns the structure of an ocean feature at several scales at once, so the vehicle can plan against it.
-
Coordinated Sampling of Dynamic Oceanographic Features with Underwater Vehicles and Drifters
Drifters go where the water goes; AUVs go where the drifters say; together they track a moving patch of ocean.
-
Towards Mixed-Initiative, Multi-Robot Field Experiments: Design, Deployment, and Lessons Learned
What actually happens when humans and several robots share a field campaign in Monterey Bay.
-
Towards Marine Bloom Trajectory Prediction for AUV Mission Planning
Predict where the bloom will be, then send the vehicle there.
-
A Robotic Sentinel for Benthic Sampling Along a Transect
The first paper: a robot that patrols a line on the seafloor and decides when to take a sample.
Thesis
-
Ph.D., Computer Science, University of Southern California, 2014
Data-driven robotic sampling of dynamic ocean phenomena — the root of everything above.