IoT Goniometer

ECE4180 Spring 2020 Final Project

Project Members: Emily Wilson, Joseph Malone, Jeremy Kwok, Kenneth Chua

Introduction

The IoT Goniometer is a device which uses a goniometer to measure a joint’s function such as an elbow or a knee, based on their degree of movement. The measurements obtained will be uploaded to the internet using the Raspberry PI Zero.

Motivation

Currently, physical/analog goniometer only take static measurements of patients for doctors. The method does not take into account the degree of movement when the patient is doing strenuous activities such as running. This project will help patients to have both their static and dynamic measurements taken while assisting doctors to automatically track their patients progress.

Demonstration

Components

The following are the components used to built this project:

Schematic

hardware architecture

Hardware Setup

hardware_setup

Software Architecture

software architecture

Main Program

NOTE

If your raspberry pi has not been set up with the correct libraries, please see how to setup the rpi.

Running the application

We are using CherryPy and Jinja2 to create a web application that can will display static, dynamic, and previous goniometer readings.

The application is broken into two parts, the goniometer class to take static and dynamic readings (in goniometer.py) and the history class to display previous static measurements (in history.py).

In app.py, we combine the goniometer and history classes, with the goniometer as the root of the application.

root = Goniometer()
root.history = HistoryPage()

We then set the application to run on the Raspberry Pi’s IP address so other computers on the same network can access the application.

cherrypy.config.update({'server.socket_host': '192.168.1.148'})

To run the application, we can just run the app.py file on the Raspberry Pi.

python3 app.py

And we can access the application by going to http://192.168.1.148:8080 in a browser on any computer connected to the same network as the Pi. Please see here for full code.

Explanation of the code

Goniometer

The goniometer can take two kinds of measurements, a static reading and a dynamic reading, which we can select from in the main measure page of the application.

static measure page

The static reading takes 100 samples and averages the readings from the flex sensor and the ADC to give one reading which is then converted to an angle using an equation from calibrating the goniometer.

for i in range(100):
      reading = adc.read_adc(0, gain=1)
      readings.append(reading)
      time.sleep(0.0125)

To calibrate the goniometer so that we can calculate an angle from the flex sensor reading, we recorded the flex sensor reading at different angles, then graphed flex sensor reading vs. angle and generated a trendline for the data.

goniometer calibration graph

The web application then redirects to the static reading display page which shows the static reading as well as gives an option to take another reading without going back to the home page.

static reading page

The dynamic reading takes 10 readings of the flex sensor then averages them for a stable reading and does this 200 times, reading the flex sensor for a total of 6.25 seconds.

for i in range(200):
      _readings = list()
      for i in range(10):
          reading = adc.read_adc(0, gain=1)
          _readings.append(reading)
      readings.append(Goniometer.get_angle(sum(_readings) / len(_readings)))
      time.sleep(0.03125)

The web application then redirects to the dynamic reading display page, which shows the maximum angle reached as well as a graph of the goniometer reading vs. time.

dynamic reading webpage

Full measurement code can be found here and web application code for the goniometer portion can be found here

History Page

The history page makes a mysql query for the 20 most recent static measurements in the database

cursor.execute('''SELECT * FROM readings ORDER BY dtg DESC LIMIT 20''')

then displays the measurements in a table.

history page

We can also use the search bar to search for records by patient name. This makes another mysql query for the 20 most recent static measurements taken on that patient.

cursor.execute('''SELECT * FROM readings WHERE name=%s ORDER BY dtg DESC''', (patient_name,));

history page with name search

References

CherryPy Team, “Tutorial 4: Submit this form,” Tutorials. [Online]. Available: https://docs.cherrypy.org/en/latest/tutorials.html#tutorial-4-submit-this-form [Accessed: 17-Apr-2020].

Maclean, M. (2020). Raspberry Pi Computing: Analog Measurement. Retrieved from https://leanpub.com/rpcanalog/read#leanpub-auto-connecting-analog-sensors-to-the-raspberry-pi

Things to Note

Full code for the history page can be found here

ece4180.github.io

IOT Goniometer