Embedded Systems
Autonomous Embedded Cat Care System
Built a real-time embedded prototype on ESP32-S3 with interrupt-driven ultrasonic sensing,
hardware timers, buffered sensor logic, and state-machine motor control for automated feeding
and cleaning behavior.
- Used ISR-based event handling and deterministic timing.
- Reduced race-condition risk with volatile variables and critical sections.
- Delivered a stable prototype that coordinated sensing and actuation reliably.
Robotics Modeling
Robotic Arm Kinematics and Trajectory Control
Created a Python workflow for 6-DOF manipulator modeling using Denavit-Hartenberg parameters,
forward kinematics, numerical inverse kinematics, and trajectory generation for pick-and-place tasks.
- Implemented SciPy-based optimization for inverse kinematics.
- Validated motion feasibility across multiple arm configurations.
- Strengthened skills in robotics control and system modeling.
Hardware Integration
Line-Tracking Robot
Designed and assembled a course robot with Arduino control, custom PCB sensing hardware,
photoresistors, motors, and a 3D-printed chassis, then iteratively tuned it through repeated track testing.
- Integrated custom sensing hardware with control logic.
- Contributed to calibration, tuning, and full-system optimization.
- Reached a measured course completion time of 48 seconds.
Sensor Subsystem
Mars Rover Sensor Subsystem
Supported Husky Robotics sensor planning for a Mars-style rover, evaluating environmental sensing
options and integration tradeoffs across gas, temperature, humidity, and subsurface measurement goals.
- Compared sensor specifications, operating ranges, and integration constraints.
- Connected sensor choices to rover science objectives and feasibility.
- Contributed to subsystem development and enclosure-oriented robotics work.
Data Pipeline
Seafood Production and Consumption Forecasting
Built a Python data workflow that ingested multi-year CSV datasets, handled incomplete records,
generated country-level visualizations, and estimated future production demand using forecasting logic.
- Structured messy real-world data for downstream analysis.
- Created country and global forecast outputs.
- Wrote targeted tests to validate core processing and calculations.
NLP / Analytics
Educational Text Analytics Workflow
Developed Python and Jupyter workflows for preprocessing educational text data, running topic-modeling
experiments, and evaluating model quality with accuracy, precision, recall, and F1 metrics.
- Worked with unstructured datasets and transformation pipelines.
- Compared machine-generated themes against expert-defined categories.
- Analyzed how modeling choices affected interpretability and subgroup performance.