Author: sdunn

This week I worked primarily on finalizing the current design of our device. I am a bit behind schedule, in that I did not finish my design planning early enough to place the part order last week. I don’t think this will pose too much of an issue, as the order will hopefully go out next Tuesday, and I can use that lead time to start writing some arduino code to interface with the controllers such that as soon as they arrive I will be able to start testing. The ECE course that helped me to develop this design is primarily 18-220 and 18-100, which gave me basic knowledge around using sensors and arduinos, supplemented by reading lots of tutorials about inertial measurement units on the internet.

Aside from working on the hardware components, I also assisted with making slides / graphics for the design presentation. Next week I hope to get the parts in, and hopefully start testing by the end of the week, but if they don’t arrive in time, I will refine the design for the foot pressure sensing unit. This is because at this time, I was focused primarily on getting the more critical back unit done(as this is somewhat a blocker for Jasmine’s work on back orientation detection until we know what data we will be able to collect).

 

Here’s a link to the sheet I used to plan parts. https://docs.google.com/spreadsheets/d/19h7SsfZl7L6Hj6JMvvSDONWY2HEp8ZFq5vFqzvN0bgU/edit?usp=sharing

 

To start off the week, I spent a good deal of time practicing and preparing for my proposal presentation. I refined my talking points to make sure it would fit into 12 minutes, as it was originally sometimes 15!

Having given the presentation, I spent the rest of the week researching sensor options. In my sensor research, I learned that you can only hook up a single MPU6500 combined gyroscope accelerometer to a Arduino at a time(there is perhaps a way to communicate multiple over the bus, but this is beyond my current analog device knowledge). With this limitation in mind, I decided to start looking at bluetooth enabled sensor units instead. This led me to two options, the MBIENTLAB MotionRL sensor for $110 per unit, and the WITMOTION sensor at about $50. While the WITMOTION sensor is cheaper, the MBIENTLAB sensor has much better documentation, and has clearly been used in developing other applications. It also includes a magnetometer, which prevents angular velocity drift. I believe this will be helpful in limiting potential data collection errors later on.  It also provides a much larger range of support than the WITMOTION sensor, allowing for low power consumption modes, and a battery life of up to 14 days. For this reason, I currently believe it is the clear choice. I am not currently behind schedule. Next week, I intend to actually order the sensors, as well as begin writing some code to communicate with the sensors so that once they arrive I can immediately begin experimenting with their functionality.

 

Link to reseach doc:

https://docs.google.com/document/d/1jEEZFMm4V-nKgTUkH4GX_SufVJgoKhpiRbq81Mextvg/edit