Embedded AI Microgrant

Image of ideation notepad and small microcontroller on table

Applications Are Open!

Program Overview

The Embedded AI Microgrant supports students in building and deploying real-world AI systems on embedded hardware, bringing machine learning off the screen and into physical devices.

Build AI-powered devices - using sensors, cameras, microphones, etc.
Receive funding & hardware - and keep what you build!
Create a portfolio-ready project - great for resumes and applications 
Apply solo or with a small team - open to all undergrad & grad students

Fall 2026 Timeline

Applications will be processed on a rolling basis while funding lasts. Priority will be given to applications submitted before September 13.

Applications Open: September 1

Priority Consideration: September 13, 11:59 PM

Final Report Submission Deadline: December 4, 11:59 PM

Optional Project Showcase: TBA

How It Works

  1. Submit your project idea
  2. Get matched to the right hardware tier and accessories
  3. Receive your hardware & funding once approved
  4. Build your project and document your process
  5. Submit your final report
  6. Keep what you build. Finish a working demo and the hardware is yours

 

Check-Ins & Community

You'll join our Discord when you're accepted. Ask questions, share progress, and swap advice with other project teams and think[box] staff throughout the project. 

Twice during the semester, we'll ask everyone to post a quick update to the group:

  • Boot-Up Check: Let everyone know that your board is booted up and everything's working.
  • Mid-way Progress Check: Post how the build is going and where you're stuck. Other students and staff can help, and you'll see what everyone else is making.

Choose your Hardware

Projects can range from simple sensors to autonomous systems. Choose the level that's right for you and your project.

HardwareBest Choice ForExample Projects

ESP32-S3

 

Sensor & audio projects  A good starting point if you're new to embedded AI
  • Basic image classification
  • Simple object detection
  • Voice or keyword recognition
  • Smart sensing and monitoring devices

Raspberry Pi 5

 

Vision & interactive projects. A good choice for many projects
  • Smart doorbell with face recognition
  • Pet-tracking robot with object detection
  • Automated medication verification
  • Interactive art with pose estimation
  • Lightweight local language models for simple interfaces

NVIDIA Jetson Orin Nano

 

Real-time & GPU-accelerated projects.
  • Real-time fall detection system for elderly care
  • Industrial quality control with object counting
  • Live video style transfer installation
  • Search and rescue drone with person detection
  • On-device language models for specialized applications

 

Which tier is right for me?

Most projects run great on the Raspberry Pi 5. It's the right call for the majority of vision, audio, and interactive builds. 

Choose the ESP32-S3 if your project is lightweight: keyword spotting, simple classification, basic sensing, or if you need a very small, low-power package like a wearable;

Choose the Jetson only if your project genuinely requires real-time GPU acceleration for tasks that a Pi 5 can't handle. Not sure? Start with the Pi 5, it does more than you'd expect.

Choose your Accessories

Choose from these accessory kits and components to support your project’s hardware needs, or propose your own Bill of Materials.

Additionally, all projects will receive $50 think[box] materials credit to purchase project materials such as 3D printer filament, wood, acrylic, adhesive vinyl, etc. 

Eligibility

Open to all currently enrolled CWRU students. No prior experience required.

Ethics Statement

This program supports the development of experimental embedded AI prototypes for educational and exploratory purposes. These projects are intended to demonstrate concepts and technical feasibility.

As with all AI technologies, participants are encouraged to consider ethical implications such as fairness, privacy, safety, transparency, and potential misuse throughout the design process.

Additional resources: