EEG to Elbow; Engineering Recovery

[This Week’s Update]

A little late but I wanted to share this once we had all had our interviews for the project. Everything will be on GitHub under our organization, and here is a link to our video that we submitted. It feels weird to be done after 4 years… Anyway:

🚀 Building a Mind-Controlled Arm: Our Journey in Neurorehabilitation

Over the past academic year, our team embarked on a bold, technically demanding project: building a wearable robotic exoskeleton arm for post-stroke rehabilitation — controlled entirely by brainwaves.

Armed with a 14-sensor EEG headset (Emotiv Epoc X) and a passion for innovation, we set out to prove that neurotechnology could be harnessed for real-world rehabilitation. 🧠 The Vision

The goal? Create a proof-of-concept exoskeleton that could interpret motor imagery brain signals and translate them into real-time physical motion — specifically bicep curls and wrist rotations. We wanted it to be safe, portable, and easy to use, while using off-the-shelf components to keep it accessible and affordable making it a suitable device for at home rehabilitation.

🔄 Adapting Through Challenges

While the original plan involved custom EEG signal processing and running everything on a Jetson Nano using ROS and Simulink, reality had other plans. Proprietary EEG restrictions, hardware instability, and time constraints forced us to pivot — often.

EEG signal processing was offloaded to Emotiv’s cloud platform.

Control logic moved from the Jetson Nano to three Arduino Unos for better stability.

Materials shifted from carbon fiber to 3D-printed PLA for affordability and adaptability.

These changes weren’t setbacks — they were smart adaptations that made the final system more robust and testable.

🔧 What We Built

Our final prototype included:

Two servo-controlled joints for elbow flexion and wrist rotation.

A custom 3D-printed frame with modular PLA components.

Safety systems like emergency stop buttons, limit switches, and current monitoring.

A distributed control architecture using Arduino boards and a Jetson Nano hub.

It was tested rigorously across electrical, mechanical, software, and user-safety systems — and passed with flying colors (minus the 90° wrist rotation, which maxed out at 45° due to time constraints).

👥 The Team Experience

This wasn’t just a technical exercise — it was a deep lesson in teamwork, resilience, and adaptability.

We faced:

Burnouts

Hardware failures

Crashes (both mental and mechanical)

Debugging marathons

Repeated iterations of 3D parts, electronic layouts, and software logic

But through it all, we supported each other and eventually were all on the same page. Regular team meetings, clear role divisions, and honest communication were our lifelines. Everyone stepped up when it mattered most!

🧪 Lessons & Highlights

Mental command training with Emotiv taught us just how personal brainwave patterns are — and how tricky it is to get clean, consistent datasets.

Hardware prototyping proved that simplicity wins when resources are tight.

Testing and validation (with 60+ structured tests!) ensured we could trust the system in real-world use.

We created a truly modular system, making future upgrades or research easy to integrate.

Team work and communication was also a huge lesson and will never be taken lightly again.

🔮 What’s Next?

There’s so much potential for this project to grow. Future work could include:

Full 90° wrist actuation

Better mental command differentiation

More comfortable, ergonomic designs

Machine learning integration for adaptive control

This is just the beginning — and we’re proud to say we’ve laid a strong foundation for EEG-controlled rehabilitation devices.

💬 Final Thoughts

This project pushed us further than we expected — technically, mentally, and as a team. But it also showed us what’s possible when you combine creativity with engineering, science with empathy, and resilience with teamwork.

We built more than just a robotic arm — we built experience, trust, and a lasting sense of pride in what we achieved.

Thanks for reading, and stay tuned for MY next update as I work on personal projects and hunt for a job now.