Mechanical Engineering Portfolio

elliotk1@outlook.comgithub.com/elliot-kaiserlinkedin.com/in/elliot-kaiser

Failure Detection System for Spring Cycle Tester

Renton Coil Spring
ESP32C++KiCadSolidWorksPiezoelectric Sensor

Note: Project details, metrics, and images shared with permission from Renton Coil Spring

Electronics Wired in 3D Printed Housing
Electronics Wired in 3D Printed Housing
KICAD Circuit Diagram
KICAD Circuit Diagram
  • Engineered an automated vibration-sensing (piezoelectric sensor) spring failure detection system, resulting in 100% spring failure detection and >99% decrease in wasted cycles per failed spring.
  • Developed custom C++ firmware on an ESP32 microcontroller to process real-time data, establishing an algorithm to distinguish normal operation from failure and enabling 24/7 automated machine monitoring.
  • Integrated a relay-driven hardware interrupt directly in series with the machine's safety loop, triggering instantaneous, autonomous machine shutdown when failure is detected.
  • Leveraged the ESP32's Wi-Fi capabilities, allowing remote monitoring on computers and mobile devices.
  • Designed a compact, 3D-printed controller housing in SolidWorks to safely isolate the microcontroller and relay.

Smart Scale & Cooking Companion

Raspberry Pi 4FlaskSupabase (PostgreSQL)SolidWorksHX711 ADC
Section View of SOLIDWORKS Assembly
Section View of SOLIDWORKS Assembly
Wired Scale with 100g Test Load
Wired Scale with 100g Test Load
  • Developed an "all-in-one" smart scale + website (Flask) for calorie/nutrition tracking from Supabase (PostgreSQL) database, weight change goals, and recipe guidance, increasing food prep efficiency by 70%.
  • Engineered a custom 3D-printed enclosure with a snap-fit locking mechanism, 0.25mm hardware clearance, and isolated cantilever load cell arm for accurate measurements.
  • Integrated a Raspberry Pi 4, HX711 ADC, OLED, 5kg cantilever load cell, and pushbuttons (I2C/GPIO), increasing accuracy with median filtering and in-app calibration.
  • Designed a guided cook flow that scrapes recipes into gram targets, walks ingredients step-by-step on hardware and website, and converts cooked yield into accurate portion macros.
  • Implemented a predictive goal-planning algorithm that estimates TDEE based on logged intake and tracks true body weight trends using a 7-day exponential moving average.

Volunteer Management Software

Next.jsReactTypeScriptSupabase (PostgreSQL)
Director View
Member View
  • Engineered a mobile application using Next.js, React, and TypeScript to centralize volunteer coordination and communication for a sailing club of 120 members, slashing overhead by over 50%.
  • Integrated a secure, real-time backend using Supabase with custom Row-Level Security policies to guarantee reliable data access across messaging and task assignment systems.
  • Built separate Director and Member views with role-based access control, enabling directors to manage schedules and assignments while members receive real-time task updates.
  • Validated efficiency gains through Keystroke-Level Model (KLM) analysis, demonstrating over 50% reduction in time-on-task compared to the club's previous manual methods.