open to research roles and internships in robotics + ml

McAllister
Blair

Data science student at UC San Diego, building at the intersection of machine learning and physical systems. I'm interested in robotics, reinforcement/imitation learning, and applying data science to human cognitive health, ultimately building things that help people.

scroll to explore

Things I've built

6DoF Robot Arm — Imitation Learning

lerobot · so101 · act policy · ps4 teleop · hugging face

imitation learning ACT policy robotics python ubuntu wandb hugging face

Built a 6 degree-of-freedom follower robot arm from hardware assembly through deployment of a trained imitation learning policy. The final system autonomously picks up a block and places it into a bowl.

  • Wrote a custom PS4 teleoperation script for collecting demonstrations
  • Recorded 50+ episodes and trained an ACT policy on the collected trajectories
  • Deployed the trained policy back onto the physical robot and validated it on real hardware

This project gave me hands-on experience debugging a full robotics stack across hardware, webcam input, controller mapping, data collection, and model training on Ubuntu.

ACT policy · training performance open in wandb ↗

Training performance for the ACT policy, trained on recorded teleoperation episodes. Loss decreased from 0.35 → 0.06. The final dataset and model used for training and evaluation are listed on my HuggingFace profile: so101-block-place-final (model) and so101_pick_and_place_v3 (dataset)

selected projects

Power Outage Severity Analysis

logistic regression · fairness audit · hypothesis testing · data science

classification fairness EDA statistics

Built a logistic regression model to predict whether a power outage would last more than 48 hours using cause, climate, and customer-related features from about 1,535 outage events.

  • model performance
    Final model achieved 0.83 AUC using log transforms and interaction features.
  • responsible evaluation
    Performed permutation testing, MAR analysis, and a fairness audit that found a recall gap between Western and Eastern regions.

Baby Schema Cuteness Prediction

computer vision · resnet18 embeddings · ridge regression · psychology

computer vision research-inspired regression python

Built a small computer vision project exploring whether infant facial features associated with “baby schema” could predict perceived cuteness. Used ResNet18 embeddings and Ridge regression on a curated image set.

  • result
    Achieved a 12% reduction in MAE over baseline using 5-fold cross-validation.
  • research connection
    Linked the modeling results to psychology literature on baby schema, cuteness perception, and caregiving response.
background

About me

I'm a third-year Data Science student at UC San Diego with a growing interest in machine learning for physical systems. During my first two years of undergrad, I focused on building a strong foundation in data science and machine learning. Now, I’m excited to apply those skills to problems that go beyond the standard scope of my major, especially in robotics and other real-world systems.

I built my first robot arm this year, from hardware assembly to a working imitation learning pipeline, with no prior robotics experience. I taught myself everything from motor control basics to training ACT policies on demonstration data.

My long term goal is to work on robots that help humans, from assistive devices to microscale medical robotics. The thread connecting my interests in robotics and cognitive science is the same: how can we use data and intelligent systems to improve human life in ways that are non-invasive, transparent, and trustworthy?

I'm currently seeking undergraduate research opportunities and internships in robotics, machine learning, and data science.

  • Imitation/Reinforcement learning & robot manipulation
  • Embodied AI & physical ML systems
  • Cognitive science & human health data
  • Assistive & medical robotics
  • Building from scratch

currently

Third year at UCSD · Data Science major · Seeking research/intern roles in Data Science · Open to collaboration

Let's connect

Open to research roles, collaborations, and conversations about robotics and ML.