'Nurse Box' is a personal vital scanner system that an individual could use at a home setting to regularly check their temperature and pulse rate with little to no effort. Nurse box records and stores your temperature and pulse rate data and even goes one step further to send you an alert when the vitals have gone past the safety levels. This device in its ideal version helps you keep important health data for analysis and lets you know when something is wrong. 'Nurse Box' is in no way a replacement to a doctor or an emergency-responder but rather, an automated nurse that checks up on you as frequently as you want and tells you whenever some vital data is off the charts.
Step 1: Original Plans and Its Evolution:
The original plan was to design the Nurse Box as described above with a fully functioning temperature-sensing an pulse rate- sensing circuit. Initially I wanted to incorporate the android application that visualizes this data as sent from the Nurse Box.
To provide some context, aside from "making" minimal circuits as required in lab classes in high school, I never really created any circuits until this semester. I was more of a programmer and felt comfortable dealing with complex code rather than simple circuits. When I signed up for 297DP with Professor Charles Malloch, I knew this was going to be out of my comfort zone where I had a high chance of failure. I tried to keep my goal project reasonably hard because I knew that if I aimed at something harder just because my peers were working on more sophisticated projects, I would end up short and disappointed. So, the original plan was nothing sophisticated on paper, but stood at the peak of a steep learning curve for me personally. I tested out numerous mini circuits trying to understand the control that an arduino had on the circuit and how I could manipulate it with code. I read up a lot about Arduino and worked through the entire SparkFun inventor's kit. It was a process but a fantastic way to truly understand circuits and code. After a few weeks of inventor's kit exploration, I embarked on my mission. I worked on the temperature circuit which took a couple of weeks to understand and code. Although direct implementations existed online, I wanted to understand and do every line of code myself. Pulse rate circuit was more challenging because it dealt with different kinds of capacitors and an IC- LM324. This complicated my circuit and I had to scale back on my plans of incorporating a bluetooth module and android app to visualize the data. It is doable and I already read a lot and learnt what I needed to do to implement this but time was limited.
Step 2: Research Done:
Research was one of the most integral and time-consuming aspects of this project and this course. I spent as many hours as I could spare understanding the otherwise abstracted concepts behind arduino and circuits. Looking at a circuit diagram and placing the components is not the hard part- it is coming up with the circuit diagram or understanding why a schematic appeared on google the way it did. Reading up about vitals and understanding what the data meant and how best to present it to the user of the Nurse Box was key in forming my goals and vision for this project. I did not find scholarly articles as helpful as simplified texts and youtube videos that broke down the flow in a circuit. Most times, after reading, I remember staring at my circuit realizing that "it works but not the way I think it works." Knowing the theory in class was very different from reading the theory searching for a way to implement it on a project. Everything that seemed as insignificant as a resistor on a schematic in textbook became significant. The concepts I learned were invaluable. I even went into a rabbit hole of wireless communication and Diffie-helman key exchange security encryption in order to implement a secure way to send data from the system to the phone. Frankly, the most exciting part about the whole project were the five hours I spent understanding cryptography, Ceasar cipher, RSA algorithm, and Diffie-Helman algorithm. However, I could not take the project to the level where I could have implemented any of these amazing new things I learned over the course of the project.
Step 3: Difficulties Encountered:
The challenges came in two parts: personal and developmental. Personal challenges involved underestimated the time commitment required for other classes that ate into the time I intended to keep for this project. Better time-management skills and a cleaner view of the big picture would have helped me finish the entirety of the project in time rather than a scaled version.
Developmental challenges were numerous. Understanding concepts took time and a lot of reading. Implementing was harder still because curve balls such as a busted breadboard and parts that had to be substituted for essential parts made exploration harder but more fun. Getting the code running was okay, but making sure it did what I thought it did involved a lot of poking around with multi-meter and the hardest aspect about it was I did not know what I was looking for. Towards the end, developmental challenges became more significant when I tried to implement a bluetooth module and it involved a revamp of the entire circuit. Another issue is the accuracy offered by Nurse Box. Temperature values are off and needs to be accounted for. If we were to get the temperature from a closed part of the body we will need to structure the circuit with a use-and throw layering for the sensor so that it is hygienic and accurate.
Step 4: Change Needed in M5:
Makerspace M5 was the ideal space for the development of this project. Going in to the first week after I had picked the project I wanted to work on, there was not much I expected to learn outside of the niche topic of vitals and circuits relating to it. However, Makerspace made it possible for me to have discussions with my peers on their projects as much as I thought about mine and in the process I believe I have learnt far more than I should have. Ayan Sengupta's project regarding pattern matching taught me so much about machine learning, training bots and the use of eigenvectors(finally!). Stephen Lendl's project regarding Weather display on mirror introduced me to Raspberry Pi and helped me realize the power of APIs and Python to get real-time updates on a system. Ben Button and I sat down to figure out transistors and how it could be used to turn a rotor by sequential current flow. I got more out of that makerspace than what I put in and it was mostly because of the culture that the place upheld where we could learn anything from anybody. I personally would not streamline anything in that space but I sure wish I capitalized on the abundance of knowledge in that room by getting them to comment and suggest ideas about my project.
Step 5: Accomplishments:
Accomplishments pertaining to Nurse Box was humbling. A temperature and pulse-sensor circuit from scratch is frankly all I can show on a demo as an accomplishment which doesn't say much. However, the process has taught me more than any text I read in the last two months. I learned a lot about human vitals, the biology behind it and the physics in measuring it. I developed the circuits from scratch and compared it to existing models and learnt in the process about the pitfalls I make while creating circuits. I finally understood the connection and use of an arduino, and how better to read datasheets for the IC involved in the project. As mentioned, I learned almost everything I know about cryptography from scratch for this class and I think right now I have the basic knowledge to develop a crude security system on Python. I even felt so motivated as to start writing an algorithm for Diffie-Hellman cryptography. Additionally, I learned a great amount about machine learning, pattern matching, raspberry Pi's and transistors from my peers during my time in the course.I also studied about Android App on Android Studio and I know I can pick up where I left on in summer to visualize the data collected.
Step 6: How to Create Your Own Nurse Box
Understanding the goal is of utmost priority. I had a huge misinterpretation of the concept of pulse and pressure which set me back by two weeks.Once the concepts are down, I would recommend starting at very rudimentary circuits if you are as unfamiliar with circuits as me. It's crucial to be familiar with the functioning of the parts and devices and the general flow of the system before you jump into the creation of a complicated circuit. Once that is done, breaking up the project into stages and working through them is going to be important. I broke the project into temperature sensor, pulse-rate sensor, bluetooth module, android app. stages. Very quickly I realized what I could hope to achieve in the limited time I had and what I could not. This helps work very realistically and keeps the whole project from seeming daunting. Blindly following steps on a website would lead to a lot of problems and extremely limited functionality. You would want to understand how your circuit behaves so you can tweak it to what you envision it to be. As for the parts of the project that are work in progress- the bluetooth module and the android app, I am going about it by working on bluetooth module setup and reception separately and then integrating this into the circuit with the necessary tweaks.
Arduino code involves a sequential flow of activating, receiving and manipulating data on loop. Once this is understood, looking up how to activate a port or receive a data is all you need to do. There is a definite logical breakdown of this and debugging the code and analyzing the circuit with multi-meters is a great way to go about it.
Step 7: Next Steps:
Nurse Box has a powerful role to play from a user standpoint. However, in its current form we limit Nurse Box's data transmission to close proximity wireless transmission via bluetooth. If we can upgrade the arduino to a Raspberry Pi, which in retrospect I should have started with, we can easily use Wifi for long term data transmission as a result of which the personal physician of the user or say, a close relative, could be kept in the loop with regard to the vitals. More vitals can be added such as pressure and this will make the Nurse Box all the more potent. Incorporating a use and throw component over the temperature sensor without compromising on the accuracy will be important if we were to take temperature from closed off body parts. Fine-tuning efficiency and accuracy and optimizing the flow of data and securing the transmission of data will be the final steps before Nurse Box can be unveiled to the user. There is definite need and space in the society for the concept of a personal vital scanner that offers as much as Nurse Box does. Lots of challenges ahead but I believe it worth trying.