From Breadboard to Breakthrough

[This Week’s Update]

A journey of transformation—turning a university prototype into a modern IoT therapy device with mobile BLE control and clinical analytics. A Video will be out soon. This only took so long as I was waiting to make the video however the construction occuring in our house has been extended by a week.

---

Introduction

What began as a humble Arduino servo controller for my Bachelors, has evolved into a somewhat professional-grade rehabilitation exoskeleton for your hand powered by an ESP32 microcontroller. Featuring wireless control, real-time analytics, and scalable infrastructure, this project showcases how engineering can elevate a concept into an (almost) clinical-ready system.

• Original Arduino Code Architecture

• New ESP Code Architecture

• Project Page

🛠️ Phase 1: Humble Beginnings with Arduino

The original concept used an Arduino Uno and a breadboard-packed setup. It controlled three micro servos via a button and offered minimal feedback through two LEDs. While functional, it had several limitations:

• Bulky and fragile

• Limited feedback and UI

• No data logging

• Poor scalability and reliability

Despite its simplicity, this version laid the foundation for all future iterations.

🚀 Phase 2: ESP32 + FreeRTOS = True Multitasking

The shift to the ESP32 was pivotal. With dual-core performance, built-in WiFi/BLE, and FreeRTOS support, this upgrade enabled:

• Concurrent BLE communication and servo control

• Real-time command processing

• Efficient power and task management

• Multithreaded servo operation and analytics logging

FreeRTOS let me isolate critical tasks—like servo actuation and communication—onto dedicated cores, improving both performance and reliability. I’ve actually become quite fond of FreeRTOS.

📱 Phase 3: BLE Mobile Control with React Native

Goodbye button mashing, hello BLE.

I developed a cross-platform mobile app using React Native and TypeScript. It connects to the ESP32 over BLE, offering:

• Session management (auto start, manual end)

• Real-time status and command feedback

• Seamless scanning and reconnect logic

• Debugging tools for error tracing

BLE commands are handled via a custom GATT service, allowing direct triggering of servo patterns or emergency stops.

🧠 Phase 4: Real-Time Data Infrastructure

Data brings insight. With Dockerized backend components, I built a full analytics pipeline:

• Mosquitto for MQTT messaging

• MariaDB for data storage

• Grafana for clinical-grade dashboards

• phpMyAdmin for DB management

This setup enables live tracking of therapy metrics, including success rates, movement quality, system health, and session history.

🧩 Phase 5: 3D Printing and Wearable Integration

To move beyond the breadboard, I created 3D-printable components:

• Hand exoskeleton (servo-fitted and adjustable)

• ESP32 mounting base

• Wearable splint integration

Printed in PLA and flexible filament, these components drastically improved form factor, durability, and comfort.

🔬 Future Pipeline: Advanced Sensor Integration

Designed for extensibility, the architecture supports future hardware:

• Pressure sensors for grip strength

• MPU-6050 IMU for motion tracking

• MAX30102 for heart rate & oximetry

These will interface via InfluxDB for high-frequency data capture, enabling machine learning-based therapy optimization in the future. These components have been ordered and should be here within a week… I hope.

⚔️ Engineering Challenges & Solutions

1. BLE Reliability: → Solved with reconnection logic, state management, and robust error handling.

2. Servo Power Issues: → Introduced staggered actuation patterns and task synchronization to reduce current spikes.

3. Real-Time Data Logging: → Implemented MQTT-based publishing with MariaDB backend and Grafana dashboards.

4. Complex Session Management: → Automated start/end logic with session-specific analytics.

5. Cross-Platform Complexity: → Used PlatformIO, Expo, and Docker Compose for manageable multi-platform development.

📈 Performance Snapshot

Feature Arduino ESP32 System Response Time ~500ms <100ms Control Interface Single button BLE mobile app Data Logging None Real-time MQTT + DB Form Factor Breadboard 3D-printed wearable Analytics Not available Clinical dashboards

🧠 Technologies Utalized

• Embedded: ESP32, FreeRTOS, NimBLE, PlatformIO

• Mobile: React Native, TypeScript, Expo, BLE

• Backend: Docker, MQTT, MariaDB, Grafana

• Hardware: 3D printing, servo integration, sensor interfacing

🧭 Final Thoughts

This project is a case study in engineering evolution—from a simple prototype to a robust, scalable, and clinically useful rehabilitation tool (in theory). It integrates best practices across hardware, firmware, mobile, and backend development.

Whether you’re building your first IoT project or scaling for real-world deployment, the lessons from this journey are universal: embrace modularity, prioritize usability, plan for analytics, and always design with the end-user in mind. I recommend you give this project a try and build it yourself. You can rund the server on any machine, just install docker or better yet use an old laptop or PI you have laying around.

Anyway I hope you enjoyed reading about my journey from a simple Arduino-based servo controller to a sophisticated ESP32-powered rehabilitation device with wireless mobile control, real-time data logging, and comprehensive clinical analytics. If you have any questions or comments, feel free to reach out!

Thanks for reading my ramble, see you next week (or two).