Graduate Teaching and Research Specialist
ECE 5760: Advanced Microcontroller Design & SOC
Masters student at Cornell University
studying Electrical and Computer
Engineering.
Immanuel bridges the gap between hardware and software. He has worked as a Systems and Embedded Engineer in the EV sector (Sun Mobility) and served as the Vice Captain of Team Assailing Falcons, an SAE Aerodesign student team ranked 3rd worldwide.
15
PROJECTS
7+
PATENTS
2
PUBLICATIONS
6+
AWARDS
Philosophy: The Art of Co-Design
"I believe the future belongs to Hardware-Software Co-Design. Real performance isn't found in code alone or in silicon alone — it lives in the handshake between them. Whether designing architectures for heavy EVs or optimizing FPGA logic, I'm driven by making atoms and bits move as one. This is my love letter to engineering: building sustainable systems where intelligence survives the real world."
Current Status
Graduate Teaching and Research Specialist
ECE 2100: Intro to Circuits for ECE
Engineering Roots
Team Assailing Falcons (SAE Aerodesign)
ECE 5160: Fast Robots Progress
Microcontroller Basics & Artemis
Setting up the Artemis board and establishing BLE communication.
Read EntrySensors and Parallel Execution
Time-of-Flight sensor integration and non-blocking parallel execution.
Read EntryMotors & Open Loop Control
Changing from manual to untethered open loop control using Artemis and dual motor drivers.
Read EntryLinear PID Control & Interpolation
Closed-loop PID controller driving the robot to 304mm from a wall using ToF feedback, extrapolation, and wind-up protection.
Read EntryOrientation Control
DMP-based yaw PID controller with derivative kick suppression, low-pass filtering, and integral anti-windup.
Read EntryKalman Filter
Physics-based Kalman Filter replacing linear extrapolation, running at 204 Hz — 7.2× faster than the ToF sensor update rate.
Read EntryStunts
4-phase hybrid state machine — KF approach, open-loop drift, PID correction, and 255 PWM return sprint. Full stunt in ~3.2 s.
Read EntryMapping
PID step-and-scan mapping using ToF sensor — 72-point polar scans transformed into a 12-segment global line map for simulation.
Read EntryGrid Localization
Bayes Filter over a 12×9×18 state grid — vectorized sensor updates rescue localization even when odometry places the prior at 6.5×10⁻¹⁵.
Read EntryReal-Robot Localization
Update-only Bayes filter on the physical robot — 18-ray ToF scans at four marked poses, with sensor_sigma tuning for real-world noise.
Read EntryInverted Pendulum / Wheelie
Balancing the RC car as an inverted pendulum with a state machine to transition from flat to wheelie.
Read Entry