Engineering

Systems Engineering Internship @ Arc Boats Link to heading

At Arc, I developed an automated high voltage test interface to validate the build quality of Arc’s 226 kwh battery.

I automated the pre-charge, pack charge, hipot and battery management system (BMS) tests by repurposing the boat’s vehicle control unit (VCU). I learned and repurposed Arc’s proprietary CAN framework to control the VCU and custom peripherals.

I also wrote a scrappy Python driver to replace $10k vendor software, which was previously impossible because of a faulty DBC. I fixed this by reverse engineering the vendor-provided CAN DBC, byte by byte. I was then able to write the driver using OpenHTF.

During high voltage battery validation tests, I caught a fatal harnessing defect + PCB design flaw leading to HVIL signal drop. I diagnosed and fixed it in 24 hours, unblocking boat production. I then remedied the flaw and improved the board in Altium.

Inductive Proximity Sensing for an Electromagnetic Prosthesis Link to heading

During my independent research role with the UCLA Anatomical Engineering Group, I worked with a grad student who developed an electromagnetic prosthetic attachment for lower limb amputations. The system involves implanting a passive ferromagnet, allowing the wound to close, and attracting it with an external electromagnet. Electromagnets are power hungry, necessitating a controls system to decide the amount of current depending on the stage of the gait cycle.

I designed and fabricated current and voltage sensing PCBs, and wrote firmware code to communicate to the analog digital converter (ADC) via SPI.

I then pulled on prior research that suggests the inductance of an electromagnet changes as a ferromagnet moves within its field. I tapped into the voltage and current of the electromagnet, mapping inductance versus distance.

I noticed some dependence on the frequency of voltage input, but at 100 hz there is a particularly noticeable trend correlating Gap Distance [mm] with Inductance [mH].

I presented these results via poster at the Socal Robotics Symposium (Irvine 2023) and the BMES Annual Meeting (Seattle 2023).

Wireless Endoscopy Link to heading

For my senior thesis, I developed the controls system for a wireless endoscopy procedure to prove that such a system could be accomplished at scale and with repeatability.

I used my custom current sensing PCB with custom C++ code. To determine whether I could feasibly implement a control system for the average female neck size, I modeled the magnetic system in JMAG (electromagnet simulation software). I researched and ordered electromagnets and motor drivers according to the neck size determined in JMAG.

I then designed a silicone mold of an esophagus in CAD, implanted force sensors in the mold, and multiplexed them via I2C to a Teensy microcontroller.

The Teensy took inputs from the force sensors and current sensor, regulating the motor driver output via a closed feedback control loop. I was able to apply consistent, repeatable force and control the system with precision, although there was slight sensor drift over time.

I won the 2024 UCLA Dean’s Prize for Excellence in Research for this project. This prize is awarded to the top 2.4% of students presenting at UCLA’s undergraduate research week for outstanding projects.

Software Engineering Internship @ Vena Vitals Link to heading

In 2022 I joined an 8 person startup (Vena Vitals) for an internship focused on non-invasive continuous blood pressure monitoring. My coursework has spanned bioinformatics, bionics, and the decoding of Utah array neural signals. One of my biggest fascinations in college has been developing non-invasive imaging techniques for the healthcare industry.

At Vena Vitals, I worked on a team of R&D engineers to conduct human sensor tests in the operating room and write algorithms identifying physiological features. My algorithms detected two important blood pressure events in the sensor capacitance data, the systolic peak and dicrotic notch.

Identifying these features was a crucial step for developing a robust mapping between the raw sensor output and corresponding ground truth blood pressure readings. This laid the groundwork for future machine learning inference.

I also conducted tests and recorded data in the operating room, observing clinical conventions and conducting clinical studies for FDA approval.