Project

# Title Team Members TA Documents Sponsor
32 Smart Pulse Oximeter
Faris Zulhazmi
Jason Machaj
Sidney Gresham
Shengyan Liu design_document1.pdf
final_paper1.pdf
other1.pdf
proposal1.pdf
# Smart Pulse Oximeter

Team Members:
- Jason Machaj (jmach5)
- Faris Zulhazmi (farisaz2)
- Sidney Gresham (sidneyg2)
# Problem

Describe the problem you want to solve and motivate the need.
The problem at hand is the inaccuracy of pulse oximeters in individuals with darker skin tones due to the way these devices interpret oxygen saturation levels. Pulse oximeters function by emitting light through the skin and measuring how much is absorbed to determine oxygen levels in the blood. However, higher concentrations of melanin absorb more light, leading to less accurate readings and potential overestimation of oxygen saturation in individuals with darker skin tones.
Addressing this problem is essential to improving equitable healthcare outcomes. A more inclusive and reliable pulse oximetry technology is needed—one that accounts for diverse skin tones and ensures accurate readings for all individuals.

# Solution

Describe your design at a high-level, how it solves the problem, and introduce the subsystems of your project.
This project aims to develop an adaptive pulse oximeter that adjusts the number of wavelengths used based on the user's skin tone. Traditional pulse oximeters often produce inaccurate readings for individuals with darker skin tones due to increased melanin absorption, which interferes with light-based oxygen saturation measurements. Many modern devices attempt to address this by using multiple wavelengths, but this approach increases power consumption.
Our solution integrates a camera and computer vision algorithms to determine skin tone and a wavelength-switching mechanism to optimize accuracy while conserving power. The device will also measure heart rate using the same optical components, making it a multifunctional health monitoring tool. All collected data will be displayed digitally for real-time user feedback.

# Solution Components

## Subsystem 1: Pulse Oximeter Subsystem

This subsystem will use infrared and red light to measure blood oxygen levels as well as heart rate. The way this works is that oxygenated blood will absorb more infrared light and pass through more red light. Deoxygenated blood does the opposite. Knowing this, we can capture and calculate the total blood oxygen level (SpO2) based on the ratio of red and infrared light passing through with a photodetector and a calibration algorithm. In order to properly measure the heart rate, the system will measure the photoplethysmography signal (PPG). When the photodetector records the light intensity, the blood volume increases as the heart beats, causing more light to be absorbed, reducing the signal. These wave-like pattern peaks correspond to the heartbeats and use the time difference between each successive peak to calculate the heart rate in BPM.

We will use the respective emitter LEDs and photodiodes:
- Red - Kingbright APT2012SECK
- Infrared - Vishay TSAL6100
- Photodetector - Hamamatsu S1223

## Subsystem 2: Color Recognition via Computer Vision Subsystem

This subsystem will utilize the “300K PIXEL USB 2.0 MINI WEBCAM” in conjunction with a flashing light to image the skin tone of the user. Using these images, color recognition will be employed to determine whether multiple wavelengths of light would need to be used to provide higher blood oxygen level measurement accuracy depending on user skin tone.

## Subsystem 3: Digital Display Subsystem

To display the contents of our measurements, data will be taken from the microcontroller and will be displayed on an external digital display. This will show the blood oxygen levels and heart rate to the user in real time.

## Subsystem 4: Power Supply Subsystem

This system must be able to operate on a rechargeable lithium-ion battery. This subsystem will provide appropriate power to each other subsystem/component using this battery with DC-DC converters (buck/boost converters). Reasonable operation time must also be available from one charge of the li-ion battery. Power efficiency can be managed via the switching of the oximeter from one to two wavelengths depending on skin tone, leading to longer operation time on one charge and higher efficiency.


# Criterion For Success

Describe high-level goals that your project needs to achieve to be effective. These goals need to be clearly testable and not subjective.

- Read blood oxygen within a 2% range.
- Read heart rate within a 2% range.
- Camera successfully captures and sends data to the microcontroller.
- Ability to change wavelengths depending on skin tone.
- Assistance via computer color recognition (to show success, try with and without to see difference in measurement)
- Correctly display measured blood oxygen levels and heart rate.

Illini Voyager

Cameron Jones, Christopher Xu

Featured Project

# Illini Voyager

Team Members:

- Christopher Xu (cyx3)

- Cameron Jones (ccj4)

# Problem

Weather balloons are commonly used to collect meteorological data, such as temperature, pressure, humidity, and wind velocity at different layers of the atmosphere. These data are key components of today’s best predictive weather models, and we rely on the constant launch of radiosondes to meet this need. Most weather balloons cannot control their altitude and direction of travel, but if they could, we would be able to collect data from specific regions of the atmosphere, avoid commercial airspaces, increase range and duration of flights by optimizing position relative to weather forecasts, and avoid pollution from constant launches. A long endurance balloon platform also uniquely enables the performance of interesting payloads, such as the detection of high energy particles over the Antarctic, in situ measurements of high-altitude weather phenomena in remote locations, and radiation testing of electronic components. Since nearly all weather balloons flown today lack the control capability to make this possible, we are presented with an interesting engineering challenge with a significant payoff.

# Solution

We aim to solve this problem through the use of an automated venting and ballast system, which can modulate the balloon’s buoyancy to achieve a target altitude. Given accurate GPS positioning and modeling of the jetstream, we can fly at certain altitudes to navigate the winds of the upper atmosphere. The venting will be performed by an actuator fixed to the neck of the balloon, and the ballast drops will consist of small, biodegradable BBs, which pose no threat to anything below the balloon. Similar existing solutions, particularly the Stanford Valbal project, have had significant success with their long endurance launches. We are seeking to improve upon their endurance by increasing longevity from a power consumption and recharging standpoint, implementing a more capable altitude control algorithm which minimizes helium and ballast expenditures, and optimizing mechanisms to increase ballast capacity. With altitude control, the balloon has access to winds going in different directions at different layers in the atmosphere, making it possible to roughly adjust its horizontal trajectory and collect data from multiple regions in one flight.

# Solution Components

## Vent Valve and Cut-down (Mechanical)

A servo actuates a valve that allows helium to exit the balloon, decreasing the lift. The valve must allow enough flow when open to slow the initial ascent of the balloon at the cruising altitude, yet create a tight seal when closed. The same servo will also be able to detach or cut down the balloon in case we need to end the flight early. A parachute will deploy under free fall.

## Ballast Dropper (Mechanical)

A small DC motor spins a wheel to drop [biodegradable BBs](https://www.amazon.com/Force-Premium-Biodegradable-Airsoft-Ammo-20/dp/B08SHJ7LWC/). As the total weight of the system decreases, the balloon will gain altitude. This mechanism must drop BBs at a consistent weight and operate for long durations without jamming or have a method of detecting the jams and running an unjamming sequence.

## Power Subsystem (Electrical)

The entire system will be powered by a few lightweight rechargeable batteries (such as 18650). A battery protection system (such as BQ294x) will have an undervoltage and overvoltage cutoff to ensure safe voltages on the cells during charge and discharge.

## Control Subsystem (Electrical)

An STM32 microcontroller will serve as our flight computer and has the responsibility for commanding actuators, collecting data, and managing communications back to our ground console. We’ll likely use an internal watchdog timer to recover from system faults. On the same board, we’ll have GPS, pressure, temperature, and humidity sensors to determine how to actuate the vent valve or ballast.

## Communication Subsystem (Electrical)

The microcontroller will communicate via serial to the satellite modem (Iridium 9603N), sending small packets back to us on the ground with a minimum frequency of once per hour. There will also be a LED beacon visible up to 5 miles at night to meet regulations. We have read through the FAA part 101 regulations and believe our system meets all requirements to enable a safe, legal, and ethical balloon flight.

## Ground Subsystem (Software)

We will maintain a web server which will receive location reports and other data packets from our balloon while it is in flight. This piece of software will also allow us to schedule commands, respond to error conditions, and adjust the control algorithm while in flight.

# Criterion For Success

We aim to launch the balloon a week before the demo date. At the demo, we will present any data collected from the launch, as well as an identical version of the avionics board showing its functionality. A quantitative goal for the balloon is to survive 24 hours in the air, collect data for that whole period, and report it back via the satellite modem.

![Block diagram](https://i.imgur.com/0yazJTu.png)