Some of these were assigned, some were class projects, and one started as a hobby that got out of hand. They cover pretty different ground: running an embedded system in perpetuity, designing something to fail on purpose, cleaning a messy multi-year dataset, and owning the CAD on a team build that got finished and donated in a semester. Together they reflect the kind of engineering work I want to be doing.
Five-year forest restoration dataset across 5 parks and ~343 hectares for the Greater Wellington Regional Council (team project).
Explore →Single-motor compliant gripper with a shear-pin mechanical fuse. Seven prototypes; reset cut from 3 min to under 1.
Explore →~$30 BOM. Running 24/7 since deployment with zero false triggers thanks to a moving-average filter built in from day one.
Explore →12×7 ft dual-power microhome built in one semester by a freshman team of 14 and donated for emergency shelter (team project).
Explore →Deep-sky imaging telescopes: setup, alignment, and calibration.
Explore →Yearly show I have been running since 2018. ~90 mortars over open water, remotely detonated; drew nearly 2,000 spectators last year.
Explore →I build and operate deep-sky imaging telescopes. The work typically requires quick setup, alignment, calibration, and troubleshooting in order to maximize imaging uptime.
Click any image to view fullscreen. Eagle Nebula, Whirlpool Galaxy, Dumbbell Nebula, Bode's Galaxy, Total Solar Eclipse, Partial Solar Eclipse, Pleiades.
I built an ESP32-based monitoring system for WPI Wireless Association's radio "Shack" to detect when the space is in use, control an ON AIR sign, and post status updates to Slack.
The system uses a lux sensor for occupancy detection, a moving-average filter to reduce false triggers, relay-controlled signage, Slack API integration, and a 3D-printed enclosure. The design needed to be cheap, reliable enough to run unattended, and simple enough for other club members to maintain.
So far, it has had no false triggers and has not gone down. During the school year, the Slack channel usually updates once or twice a day; over the summer, it slows to roughly once a week.
Wellington's regional council had years of forest restoration data spread across five parks and wanted to know what it actually said. Were the plantings surviving? Could the data show how the program was progressing, and help justify future investment?
My team had six months to work within a program that had already been running for 5–6 years. The scope covered hundreds of planting sites, roughly 343 hectares in total, and contractor survival audits that varied heavily in quality.
I cleaned and organized multi-year contractor survival audit data, then used it to build an ArcGIS map covering the five regional parks. The map included park, planting area, and monitoring site layers, with survival and mortality metrics, custom survival data pop-ups, and attribute tables to make the data easier to interpret.
I also researched possible next steps for future monitoring, including drone imagery, LiDAR canopy-height estimation, photogrammetry, and AI-assisted stem counting, along with the carbon credit and financing background tied to the restoration program.
By the end, the council had a cleaner dataset, a map that made the survival data easier to understand, and a better idea of what remote-sensing tools might make sense as the project progresses.
Led the mechanical design of a compliant gripper for an IEEE robotics competition.
How do you pick up and put down a game piece with only one motor? I used a compliant gripper that held the piece securely, then released it when the motor drove through the drop-off motion and broke a pencil-lead shear pin. The shear pin acted as a mechanical fuse, letting the gripper fail predictably instead of damaging the whole assembly.
The gripper in action, the compliant gripping mechanism, and the shear-pin collar and retainer.
What started as a small backyard fireworks show in 2018 has slowly turned into something the whole neighborhood looks forward to. Last year, close to 2,000 people attended. As it has grown, I have had to take the safety and planning side much more seriously.
The show usually uses around 90 mortars, and another 15 to 30 cakes. The runtime changes year to year. The shortest show was about 10 minutes, and the longest was a little over 25.
The firing platform is set up over open water, which gives a natural standoff distance, a safe place for debris to land, and allows more people to watch from their road end docks, or front yards. Firing is handled remotely, so nobody is near the platform during the show. I also clear the date (typically Independence Day) with the local police chief beforehand.
Everything I use is consumer-grade, so it stays within normal local rules and does not require pyrotechnician licensing or special permits.
For me, the best part is picking the shells, building the sequence, and obsessing over pacing. It is still my favorite way to close out the summer.
The 2021 floods in Waverly, Tennessee displaced a lot of families, and Second Chance Ministries was trying to put together emergency shelter for them.
Built in one semester, with a team of fourteen freshman mechanical engineers. It had to run on either grid power or solar, and the design needed to accommodate: mothers with infants, homeless students, someone in a wheelchair, or someone with limited vision.
I was the only CAD modeler on the project, so every dimension and how the parts came together ran through my work. The build itself involved trig for the truss angles, procurement for non-donated parts, and dual-power wiring for its solar and grid modes.
Finished in one semester, delivered, and donated to Second Chance Ministries. Lipscomb's news covered the project.
Faculty: Dr. David Elrod and Dr. Kirsten Dodson, Lipscomb University, Raymond B. Jones College of Engineering. Donated to Second Chance Ministries, founded by Shirley James.