Capstone is a fundamental requirement for ABET accreditation and graduation. At OSU, ECE capstone is a two class sequence:
- ECE 3906 which is focused on project management, research, and design. (Typically fall semester)
- ECE 4905 which is focused on implementation, iteration, and validation. (typically spring semester)
I am the lead mentor for 3 sections of capstone that are taught by Dr. Saideh Zia, who is the director of capstone for the ECE department. Mentoring is similar to acting as a Teaching Assistant for the course, however due to the nature of capstone projects, I act more as a technical advisor. In 2025, I was promoted to a Lead mentor. My responsibilities include:
- Create capstone projects
- Collaborate and coordinate with sponsors to create and manage capstone projects
- Guide students in technical decision making
- Provide feedback and advice to students on technical and behavioral aspects
- Resolve conflicts when they arise
- Lead and organize the other 5+ mentors
- Make resources and reference materials for students
- Meet with students during the 8 hours of class/lab plus out-of-classroom meetings
- Deliver lectures and run classroom activities
- Evaluate presentations and design reviews (PDR, CDR, FDR).
Being a capstone mentor is unique from other TA positions since there is little grading and most of the time is spent nerding out and sharing knowledge/experience. It is also an extremely fulfilling experience to see the projects come to life and watch students grow their skills.
Note: since I am actively working this position, some of this information may be out of date and some new information may not be added.
Making Capstones
Creating capstones is a delicate task of balancing difficulty, realism, and intrigue. Another challenge is that when creating capstones is that we do not know the capabilities of students that will be assigned the project. Creating in-house capstones that are faculty sponsored is fairly straight forward since communication is quick and all we are doing is establishing a project description, scope of work, and technical requirements.
Creating Industry sponsored capstones is far more difficult. We try to offer industry sponsored capstones for every student who wants it. Until 2 years ago, ECE capstone had zero industry sponsored capstones, and through Dr. Zia and the mentors’ efforts, nearly every capstone in her sections are industry sponsored. While the process for creating these capstones vary a little more: the general process is:
- Reach out to companies that have prior OSU affiliation or alumni
- Establish a both a technical and an administrative contact
- Discuss project ideas
- Negotiate a project contract (scope of work and funding)
Projects I Created
More details will be added later.
Peak Surfer
A project made in collaboration with NSA (National Security Agency) and sponsored by the DIU (Defense Innovation Unit). The objective is centered around secure jam-resistant wireless communication.
RISC-V CPU Design
Given OSU’s abysmal computer engineering curriculum, this project encourages students to explore modern processor design on their own. The goal is less about tape-out and more about optimal design, exploring microarchitectures, and design verification.
FPGA Motion Profiling
This project is to create a hardware accelerated computer vision pipeline for ultra-high speed and ultra-low-latency real-time motion profiling for industrial robotics.
Notable Projects
During my time as a mentor, I’ve worked with a lot of memorable groups. Some of them impressive in how great they were, while others…
Here’s a list of some projects that I believe were technically impressive:
Lithostepper
Another capstone mentor who is a PhD candidate started a Hackerfab chapter at OSU and sponsored a capstone to make a lithostepper with a budget of $5k. The HackerFab project, started in CMU, intends to create affordable, accessible, and open source semiconductor manufacturing technology. The OSU HackerFab chapter had already assembled a maskless photolithography device with a manual XY stage. The capstone group’s objective was to automate the movement of the stages to sub-micron precision so that designs could be tiled.
To accomplish this level of precision, and to allow the tiling of designs on a single wafer, the group designed both a micromovement system and a wafer localization system. The wafer localization took advantage of the existing assembly by mounting a camera to the optical viewport that was typically used for manual alignment. We used OpenCV an edge detection system that could localize the corner of a wafer as an origin and then detect the corners of each etch so up to four separate projections could be tiled on a single wafer allowing for larger die area. The micromovement system was driven by a piezoelectric stepper that could move the stage exactly 80nm a step. This control loop was closed using an inductive absolute linear encoder in each axis. To keep costs low, all components were off the shelf and all of the integration work was done manually. This meant all integration work was manual; e.g. the encoder heads required custom packaging and mounting. All custom components were 3D printed to maintain accessibility.
Despite being an ECE project, there was a ton of mechanical and metrological work in this capstone. At sub micron scales, even metal becomes unstable since thermal expansion starts causing issues. This group was a pleasure to work with and they were unafraid to take on new challenges (unlike many groups unfortunately).
Other projects
The following is a list of projects I’ve had a large part in mentoring. This is a not a comprehensive list, but covers the groups I was most involved in. This list omits the projects I directly created and the notable projects above.
- High bandwidth X-Ray communication system. Sponsored by ESL
- Self driving car SIL (Software In the Loop) and HIL (Hardware In the Loop) simulation. Sponsored by GM and Buckeye Autodrive SAE team.
- Generalizable HMI (Human Machine Interface) for self driving cars. Sponsored by GM and Buckeye Autodrive SAE team
- LiDar/mmWave people and activity perception system. Sponsored by Honda.
- Real-time OCR (Optical Character Recognition). Sponsored by Honda.
- 3-axis sub 30nm precision micro-positioning system for chip die inspection. Sponsored by Intel.
- AI powered line judge for Tennis implemented in a smartphone app. Sponsored by Qualcomm.
- Multi-camera 3D motion reconstruction system. Sponsored by Artificially Intelligent Manufacturing Systems (AIMS) Lab.
- Long-life mmWave based occupancy and vehicle detection. Sponsored by The Genie Company.
- VR gloves with finger tracking, haptic feedback, and temperature control. Sponsored by the Engineering Education Department.
- UAV control, path planning, and dynamic obstacle avoidance. Sponsored by MathWorks.
- Sensor-less motor position control using onboard tinyML inference. Sponsored by Texas Insturments.
- Remote HMI for autonomous heavy load mining vehicles. Sponsored by Caterpillar.
- On device sign language transcription on AR glasses. Sponsored by Qualcomm.
- Frequency Selective Surface (FSS) and simulations for RAM (Radar Absorbing Material). Sponsored by GE AeroSpace