Iteration 4 Dominik - IOT and Cloud putting it all together.

 So far the functions of the Steam Punk  Hat are:

 

Accelerometer tracking movement

Gyroscope tracking the tilt of the head

Tunnelling Breath to the microphone to create a sensor measuring breath in and out

Pulse meter measuring the BPM 

The hat has a bellows with an actuator servo moving up and down.

 

 

It’s now time to put it all together and have this data be sent to a cloud dashboard where decisions could be made based on the certain data values. This project now enters the smart wearable tech sector. Wearable devices are becoming more and more popular. Lots of people are wearing smart watches every day but there is research being done today technologies like wearable clothing. This project fits nicely here as we have a lot of sensors that give a lot of information about the human body for example:

 

Movement: counting the steps and how fast the person is going

 

Pulse : The heart rate of the person, are they stressed, maybe low pressure? Maybe even absence of pulse? 

 

Breath : Physical and mental state of a person can be deduced from the way they breathe. (https://ideas.ted.com/feeling-anxious-the-way-you-breathe-could-be-adding-to-it/). A lot of research is done in this field and it relates to how you breathe and also your posture. With a correct breathing technique you are guaranteed better sleep, reduction of anxiety and stress and have greater control over your state of mind. Will hoff very famous for his super human ability to withstand extreme cold temperatures says his secret is the breathing. https://www.wimhofmethod.com/breathing-exercises.

 

The Position of the head especially the tilt is also a good indicator of posture which also has an impact on the wellbeing of a person. Bad posture is also a indicator of poor physical condition of exhaustion.

 

 

From the above sensors many observations can be taken the most interesting to us are:

 

·      Heart – BPM

·      Breath – In, out and hold

·      Steps – Step counter, step per minute, toggle step, moving yes or no.

·      Posture – is head down, raw tilt value

·      Fall indication- on free fall trigger

 

 

 

 

Once its known what to measure and take from the sensor the data can be sent to the cloud. We used MQTT and the microbit cloud expansion board. After setting up a beebotte channel and dashboard for the project and setting it up in the microbit code it was time to start sending data. The beebotte dashboard is not the best for real time data so an app like MQTT IOT panel can be used and set up by creating a panel and adding the account keys as username and password to the app. 

 

The first test was to see if it’s possible to receive live data from the pulse sensor on a graph and having a little monitor on the phone. Measuring the time it takes to send an MQTT message is only 4ms it was possible but the graph quickly got messy as the sensor is sensitive to the pressure and it would need to be scaled properly before it is sent. 

 

Since the amount of data being sent is a lot of individual values having something in real time is not viable and also the with the new design pattern we want all the code on a single microbit inside the hat. 

 

While sending the data onto the Beebotte broker we found a few interesting bugs or issues. 

 

1.     Sending the data to the broker over a certain limit of characters is not cleared by the esp chip on the cloud board and its appended to the next message.

2.     Having the text log on the dashboard actually buffers the messages in a queue until they are printed this slows the retrieval of the data down.

3.     The cloud board needs to be powered of the battery not the usb supply as the amount of current is not enough for the internet chip to perform decent comms.

 

Here are some screenshots of how this looks like: the first one shows the heart rate live the second shows all the fields on the dashboard.









 

My ideas for using this data that’s in the cloud is:

playing a song that matches the steps per minute or BPM of a person,

Gamifying of breathing patterns and techniques such as Will Hoff’s method,

Sending a  message to a trusted contact when there is a very low or high heartrate and fast or irregular breathing.

Sending someone an animated live beating heart matching persons real heartrate.



The code :

https://makecode.microbit.org/_DmP5CXaARLRF



The Next Blog iteration will close out the project and that's where I will explain how to set up the Project and explain how the code works.


Dominik Wawak, April 2023.


Comments

Post a Comment

Popular posts from this blog

Iteration 5 Dominik - Finished Product

Image
  Closing out This is the final entry to this blog, In the last blog I explained some applications and the Internet of Things feature of the Steam Punk Hat as well as all the features we had on it.  In this blog I want to pass over the work as well as share a final demo of the product we created.  To summarise in 10 weeks we went through 5 iterations: Mapping movement,  Working with heart rate sensors, Breath Mapping, Cloud dashboard design and feeding everything in. Speaker and sounds from a dfPlayer Mini module. Every iteration had some creative elements apart from the coding, A lot of brainstorming and designing even 3D printing. The mapping of these sensors was mapped to hardware such as buzzer sounds, lights and servos. The main challenge was to make a stable and reliable mapping.  A new addition to this iteration was this DFPlayer Mini board that allows to play sounds from an sd card and comes with a microbit extension. What do we have in the end?  A ...
Read more

Iteration 4/5 Jack - Improving Sensor Values readouts and Finalising.

Image
Introduction: This iteration involved updating the code and improving the outputs for both the Pulse Sensor and Breathing iterations of the project. This iteration is also the final one where everything is being closed off. I was responsible for upgrading and improving the functionality of the sensors while Dominik worked on sorting out the cloud connectivity with MQTT and the hardware side of the hat. Changes: The sensors need to be adjusted to work better. Some general adjustments were made to the systems to function more consistently. These changes ranged from adjustments to values like thresholds/slopes or even adding flags and triggers to the system. I also needed to add a few more value outputs to the system, including: Pulse Sensor BPM Breathing In Time Breathing Hold Time Breathing Out Time Breaths per Minute  Pulse Sensor BPM: The BPM of the pulse sensor was pretty easy to calculate. The system already counted the pulses so I just had to count every second one. This was be...
Read more

Iteration 1 Dominik - Mapping movement speed to train sounds played on a phone.

Image
Steam Punk theme: Hardware:  Microbit, steam punk costume  Software: This iteration uses the microbit bluetooth HID extension from make code extensions.  https://github.com/bsiever/microbit-pxt-blehid . It is straight forward to use and provides many options on controlling a bluetooth device by sending it bluetooth codes some which are common to all devices like play - pause - skip songs. The thing to keep in mind when using this library is to disable the radio functionality of the microbit.  I used a serial window on Mac by first listing my usb devices  /dev/cu.*   and later using this command to start a serial terminal  screen /dev/cu.usbmodem1302 115200.  On Windows we used a tool called  https://os.mbed.com/users/borislav/notebook/serial-port-plotter/  which is very good for plotting a graph. We used this to calculate the different accelerometer values for when a person is running or walking. This microbit project creates a 3D a...
Read more