Week 9 Update

This week brings good news and very bad news, the good news is that the code actually works. We edited to code to show a hyponogram of the users sleep cycle. The thresholds aren't exact as everyone's movement while sleeping can vary by a degree. The bad news is that we cannot get the device to amplify our  eye movement accurately due to the 60Hz interference that are generated by the alternating current in wall sockets. The device will work well if there is no outside interfenrence, but a place to sleep with no interference is hard to find. We also realized that tracking sleep cycles can get really tricky with EOG and accelerometers as N2 and N3 sleep are very similar with respect to eye and muscle movement.

We spent a lot of time this week talking about what we would chance about the project if we had more time. The main thing that we would change is to entirely redesign the circuit board. At the moment the combination of the circuit board and the DAQ is very awkward to have to sleep with. We also want to move from MATLAB to a mobile device such as an Android or iOS phone to act as the data acquisition unit along with a replacement for MATLAB. The electrodes also had a difficult time staying attached all night with alligator clips attached to them, we found snap connector electrodes that are also reusable. The last thing that we would want to do is acquire a polysomnograph to form more accurate thresholds for the different forms of sleep.

Week 8 Update

This has been a great week in terms of progress for the group. We tested the electrodes in lab on the oscilloscope and the readings were perfectly sinusoidal. It was one of the only amplifiers in the class that actually worked exceptionally well. Although we found out that we made a mistake when building the amplifier circuit. We didn't realize that when the product description said that the amplifier integrated circuit contained two amplifiers it meant it could handle four electrodes while we are only using two. What this means is that we built twice as much as we actually needed. In reality this isn't completely bad since we would have had another way to test the electrodes if one side didn't work. We also started testing the electrodes using the mask, which worked surprisingly well without any of the modifications made yet. The mask helped to keep the electrodes in place even when there were alligator clips connected to them. As the week progresses we will be testing the accelerometer and electrodes together.

This video shows what the electrode signal looks like when connected to the oscilloscope. This test was conducted with one electrode next to each eye.

This is what it looks like when the electrodes are connected. We later found out that we only need one alligator clip on each side instead of two. 

Week 7 Update

This week has been extremely productive for us as we have gotten the mask in and started testing everything. We focused on the accelerometer first, Michael has finished the test program for the accelerometer while Alex and Julian made the wristband to hold the accelerometer. Julian has also downloaded the top rated accelerometer app on his iPhone in order to test both the accelerometer app and our device at the same time. In class we were able to test the to see if the amplifier worked, and to our surprise we found out that we didn't burn out the amplifier while building it. We were able to use an Oscilloscope to see the readings from test electrodes placed on Julian. In the coming week we are going to fabricate the mask and learn how to interface the electrodes with MATLAB in order to test the entire device at once.

The accelerometer wristband. 

An idea we were toying around with for the future. To attach the Data Acquisition Unit to the user with either its own wristband or with the accelerometer.

Julian and I testing the amplifier and electrodes. 

Week 6 - Update 2

We've spent a lot of the week thinking about how we want the output to look. As you can see in our previous post we have made a flowchart of how we think that we want the data to be logged and processed. This flowchart is just a preliminary one so it doesn't go into any detail about how the program itself will work. We also spent time this week comparing other programs that use an accelerometer to track sleep to see how accurate they really can be.
In other news we have been thinking about how we should attach the accelerometer to the person that is sleeping and have come up with the idea to make a wristband with a small pouch that can hold the device. The wristbands were actually delivered already so now we just need to cut up one of them and sew it to the other to make a pouch. Also, the EOG sensors have finally arrived! We are excited to start testing them to see how they work and how we can implement them.

The two wristbands to hold the accelerometer in place.

99 EOG sensors, we tried one on to see how it feels. 

Week 6 - Update 1

The preliminary flowchart on how we want the program to output data. 

We have made an initial flowchart of how the data will be collected. I figured it would be good to put our thought process at the moment on paper.

Week 5 Update

We have encountered two problems over the past week, but we are still on schedule for the most part.  The first problem we have run into is that the mask has been delivered for over a week now but we haven't gotten an email telling us to pick up our package. The second problem we have run into is that the electrodes never actually shipped. We called them up a few days ago wondering where the package was only to find out that they never shipped it by accident. They said that they would ship them overnight, so we should receive them soon. Schedule wise we are actually exactly where we want to be. We will be done with the electronic section this week and we will start on the mask portion this week.

Even with these problems we have still accomplished. We have made all the alligator clips necessary for the project, we have grounded the entire circuit, and we are still working on the code that ties everything together. We have the accelerometer working in MatLab, yet we are trying to filter it to get it to be more accurate than it is right now. We hope that by the end of Week 6 the accelerometer can be finished and be ready for testing while we put together the EOG portion of the circuit.

Now that all the parts have been delivered I can give a breakdown of how much we spent on the entire project.

• Electrodes - $15
• Mask - $10
• Circuit board - $5
• Batteries - $10
• Amplifiers - $10
• Wire/Alligator Clips/Resistors/Capacitors - Free
• Total - $50

Week 4 Update

The group has gotten a variety of things done this week, yet we are still somewhat behind where we really wanted to be by now. The circuit itself is just about finished, we are still waiting on the EOG sensors that were ordered a little over a week ago. Without the sensors we aren’t going to be able to test as much as we would like to. Although, the accelerometer is somewhat interfaced with Matlab so we can begin testing with the accelerometer. All of the group members at this time are actively learning how to use Matlab, since only one of us knows how to use it right now. We are also talking about how much data we actually want to collect. An idea that is floating around right now is to not collect data from the accelerometer while the person is in REM sleep in order to organize data more efficiently. All of the group members have been conducting active research on the topic that they were assigned such as what will affect sleep habits, and why will this product be better than existing solutions.

We have added Frequently Asked Questions to its own tab, both required questions and questions that we have been getting asked from everyone we've explained our project to. We have also updated our project proposal to reflect our most recent revision of it. 

The Accelerometer and Data Acquisition Unit

The rear of the circuit board with a proposed grounding wire.

The front of the circuit board with all the connections made.

Week 3 Update

Over the past week the group has made a large amount of progress regarding out design. So far we have acquired all the parts to our circuit board, the accelerometer, and the Data Acquisition board. We have ordered the sensors and the mask, which are the only two components that we don't have yet. Everything has been soldered onto our circuit board, but connections have not been made yet. We are going to wait until the next lab section to make sure that all connections are made properly. The program that tracks the accelerometer's movement in MatLab, which is being made by Michael, is close to being completed while other group members are looking for a way to be able to log this data. All of the members are still conducting research on a wide variety of topics such as the different forms of sleep, how accurately we can track the different states, and how to use Matlab to track this data.

The preliminary circuit board. At this point everything but the grounding is connected. 

The schematic for the amplifier circuit

The group was given a very specific way to build the circuit board pictured above using the schematic also pictured above. The signal from each of the electrode leads connected to the user's face will be passed through a resistor-capacitor high-pass filter. This high-pass filter will be used to reduce the noise of the outputted signal. There will also be one resistor connected to each side of the amplifier which sets the gain of the signal to 5000x (there is a typo in the schematic).


Edit: My apologies for getting this post out late, it was written on 4/19/13 but I somehow hit save as draft instead of publish.

Week 2 - First Post

Welcome to Group 3's Blog in Engineering 103-094! The purpose of this blog is to keep you informed on the progress of our design, a device that monitors your sleep habits. We will try to post as much as possible on here, but we will post at least once a week. 

We have a design in mind right now to be able to track what type of sleep you are in and log it to a computer. With this information the computer should be able to give you a detailed view of your sleep. This information can be used to help make your sleep better or bring to a medical professional to give advice on how to make your sleeping habits better.