Theory & Background

EMG (electromyography) measures electrical signals generated by muscle contractions. MyoWare sensors amplify and rectify the EMG signal to a 0–3.3V analog output. Baseline (resting) EMG is noisy but low (~0.1–0.3V). Strong muscle contraction: 0.8–2.5V. Threshold detection: if EMG > 0.6V for >100ms → detect voluntary muscle activation. Bicep EMG → intent to flex elbow. Tricep EMG → intent to extend. Map to exoskeleton joint command.

Components & Requirements

10 components required for this project.

#	Component	Purpose	Qty
1	MyoWare EMG Sensor	Muscle activity detection on arm	x4
2	High-Torque Servo (DS3218, 20kg⋅cm)	Joint actuation	x4
3	3D-Printed Orthotic Shells	Custom-fit structural support	x1
4	IMU (MPU6050) ×2	Joint angle measurement	x2
5	Arduino Mega	Real-time EMG processing and control	x1
6	Velcro and Padding Kit	Patient comfort and safety	x1
7	Lithium Battery (7.4V 5Ah)	Portable operation	x1
8	Force Sensitive Resistors	Detecting patient grip and contact force	x4
9	LCD display (patient feedback)	Exercise progress and alerts	x1
10	Emergency stop button (patient-accessible)	Immediate safety cutoff	x1

Step-by-Step Implementation

Follow these 3 steps carefully.

1

EMG Signal Processing for Intent Detection

EMG (electromyography) measures electrical signals generated by muscle contractions. MyoWare sensors amplify and rectify the EMG signal to a 0–3.3V analog output. Baseline (resting) EMG is noisy but low (~0.1–0.3V). Strong muscle contraction: 0.8–2.5V. Threshold detection: if EMG > 0.6V for >100ms → detect voluntary muscle activation. Bicep EMG → intent to flex elbow. Tricep EMG → intent to extend. Map to exoskeleton joint command.

2

Safety Systems for Medical Device

Safety is paramount for rehabilitation devices: maximum joint angle limits (hardware hard stops + software limits), maximum force limits (FSR feedback stops motor if contact force excessive), emergency stop (patient-accessible button immediately removes power to all servos), velocity limits (joints cannot move faster than 30°/s to prevent injury), alert monitoring (battery low, servo overtemperature), and session time limit (auto-stop after 30 minutes of continuous use).

3

Rehabilitation Exercise Modes

Implement three therapy modes: Passive (exoskeleton guides arm through full range of motion, no patient effort required — for acute stroke phase), Active-Assisted (patient provides partial force, exoskeleton provides remainder — most common), Active-Resistive (patient moves against exoskeleton resistance — for strengthening phase). Display repetition count, range of motion achieved, and force applied on patient screen. Log all session data for therapist review.

Code & Implementation

Core code for exoskeleton.ino:

exoskeleton.ino
  C/C++

// EMG-controlled elbow joint
#define EMG_BICEP  A0
#define EMG_TRICEP A1
#define ELBOW_SERVO 9
#define MAX_ANGLE 145
#define MIN_ANGLE 20

int current_angle = 90;
Servo elbow;

float readEMG(int pin) {
  float sum = 0;
  for(int i=0; i<50; i++) { sum += analogRead(pin); delayMicroseconds(200); }
  return sum / 50 / 1023.0 * 3.3; // Averaged voltage
}

void loop() {
  float bicep  = readEMG(EMG_BICEP);
  float tricep = readEMG(EMG_TRICEP);
  
  // Proportional assist control
  if(bicep > 0.5) {
    int assist = map(bicep*1000, 500, 2500, 0, 3); // 0-3 degrees per loop
    current_angle = min(current_angle + assist, MAX_ANGLE);
  } else if(tricep > 0.5) {
    int assist = map(tricep*1000, 500, 2500, 0, 3);
    current_angle = max(current_angle - assist, MIN_ANGLE);
  }
  elbow.write(current_angle);
  delay(20); // 50Hz control rate
}

Testing & Troubleshooting

Test Exoskeleton for Rehabilitation by verifying each subsystem individually before full integration.

!

Troubleshooting Tips

Verify power voltages, check ground connections, use serial monitor for debug.

Real-World Applications

*Post-stroke upper limb rehabilitation

*Spinal cord injury motor recovery

*Cerebral palsy therapy assistance

*Post-surgical recovery exercise

*Parkinson's tremor compensation research

*Industrial worker fatigue assistance

*Military load-carrying augmentation research

*Prosthetic limb interface research

Extensions & Next Steps

Add haptic feedback glove for sensory restoration
Implement biofeedback visualization using VR headset for motivation
Build bilateral therapy system (mirroring healthy arm to guide affected arm)
Add brain-computer interface (EEG) for higher-level intent detection
Develop pediatric version for children with developmental disabilities

Interactive Playground

Coming Soon

An interactive simulator will be available here — simulate circuits and run code in-browser without hardware.

Frequently Asked Questions

How effective is robotic exoskeleton therapy compared to conventional physical therapy?

Meta-analyses show robotic upper-limb rehabilitation produces statistically significant improvement in motor function (Fugl-Meyer scale) for stroke patients, comparable to dose-matched conventional therapy. Key advantage: robots can provide precisely controlled, repetitive movements at high dose (hundreds of repetitions per session vs 30–50 manual), which is critical for neuroplasticity. The Lokomat (lower limb) and InMotion (upper limb) are FDA-cleared clinical devices. Home-use exoskeletons enable continuation of therapy between clinical sessions.

Introduction