TIIC 2016 North America: Powered Programmable Elbow Orthosis


University: Texas A&M University
Team Members: Nathan Glaser, Joe Loredo, Rafael Salas, David Cuevas
TI Parts Used:

Project Description

Problem Statement

The elbow is one of the most active joints on the human body. However, structural damage and

neurological muscle disorders can impair this essential joint. Everyday tasks, like drinking a glass

of water, become difficult.

Goals

Our goal is to create a programmable elbow orthosis to

1. Provide structural support for the elbow

2. Resist and restrict harmful ranges of motion

3. Assist with intended elbow movements

Features
  • Assists disabled individuals with intended elbow movements
  • Android app allows the user to personalize the orthosis
  • A kill switch and pressure sensors allow for increased user safety

Resources

All files included in submitted zip file, including git hub link and PCB schematics.

User's Guide

The powered, programmable elbow orthosis consists of a unifying mechanical structure with five embedded electrical subsystems.  The mechanical structure supports the afflicted joint and mounts the majority of the subsystems. Each of the five electrical subsystems connects with others.  Specifically, the power subsystem consists of a rechargeable DC battery which distributes power to each of the subsystems.  Electronic circuitry regulates this power flow to the processor components (producing 3.3 V for the microcontrollers, sensors, and displays) and electromechanical hardware (providing 12 V for the motor).  The body-to-sensor interface captures electric muscle activity through an array of conductive electrodes.  It then reduces signal noise through an analog bandpass filter with cutoff frequencies at 10 Hz and 300 Hz, the accepted range for electric muscle signals.  A dedicated electromyography (EMG) analog-to-digital (ADC) converter, namely the ADS1299, converts these measured analog voltages into a digital value.  A digital high pass filter then removes the DC drift that results from the changing impedance of the electrodes.  The central processor uses these filtered voltage readings to distinguish muscle activity from inactivity, combines this information with user specifications, decides how to act, and finally transmits a pulse width modulated (PWM) signal to the motor drive.  The motor-to-body interface receives the incoming PWM signal and passes it through a DRV8848 motor driver which actuates two DC brushed motors.  The motor subsequently articulates the elbow with roughly 1 N • m, the torque required to lift a glass of water.  A potentiometer provides feedback for the motor.  Finally, the user interface allows for dynamic calibration of the sensors and motor, physician customization, and diagnostic data display.

Project Overview: