Heart Rate and SpO2 Measurement System

Embedded AI-enabled vital-sign monitoring pipeline for real-time heart rate and SpO2 estimation on Arduino Nano using MAX30102 photoplethysmography (PPG) sensing.

Overview

This project implements an end-to-end edge system that acquires optical biosignals, computes heart rate and oxygen saturation, validates signal quality, and renders metrics on a local display with no cloud dependency.

It is designed for practical prototyping of low-cost health-monitoring devices where latency, simplicity, and offline operation matter.

Primary engineering goals:

• Reliable PPG signal acquisition from MAX30102 (red + IR channels)
• Deterministic on-device inference using a fixed-size sample window
• User feedback and fail-safe behavior when measurement conditions are invalid

Real-world use cases:

• At-home wellness monitoring prototypes
• Education and lab demonstrations for biomedical signal processing
• Baseline reference implementation for wearable or bedside IoT productization

Architecture

The system follows a streaming edge-compute architecture:

Input sensing -> Signal buffering -> Algorithmic estimation -> Output rendering

Component breakdown:

• Input Layer: MAX30102 captures red/IR light absorption data over I2C
• Processing Layer: firmware maintains 100-sample buffers for red and IR channels
• AI/Algorithm Layer: SpO2 and heart-rate estimation via Maxim algorithm routine
• Validation Layer: confidence flags gate output publication (validHeartRate, validSPO2)
• Presentation Layer: 16x2 LCD displays SPO2/HR or No Finger state

Design decisions:

• Edge-only computation avoids network latency and privacy exposure
• Fixed buffer length enables predictable memory and timing on microcontrollers
• Finger-detection threshold prevents false readings during low-contact conditions

Architecture Diagram

Diagram description:

1. Sensor Node: MAX30102 acquires Red + IR PPG stream
2. Transport: I2C data transfer to Arduino Nano
3. Buffering Block: rolling arrays of 100 samples per channel
4. Estimation Block: maxim_heart_rate_and_oxygen_saturation()
5. Validation Gate: validHeartRate AND validSPO2 checks
6. UI Output: LCD prints SPO2 and HR, or No Finger fallback state

Features

• Real-time heart-rate and SpO2 measurement on embedded hardware
• Finger-presence detection to reduce invalid output states
• Deterministic batch estimation from synchronized red/IR samples
• Local LCD telemetry for standalone operation
• Hardware-friendly implementation for low-power microcontroller platforms

Technical Highlights

• Signal quality gating: display updates only occur when both validity flags are true
• Embedded pipeline design: acquisition, inference, and rendering are coordinated in one control loop
• Throughput-oriented sensor setup: high sample rate and averaging tuned for robust capture
• Resource-aware implementation: static buffers and compact data types for constrained RAM
• Failure-path handling: explicit No Finger state improves operational safety and usability

Tech Stack

• Hardware: Arduino Nano, MAX30102 sensor, 16x2 LCD, 10K potentiometer
• Firmware: Arduino C/C++
• Communication: I2C (Wire)
• Libraries:
  • MAX30105 sensor driver
  • spo2_algorithm (Maxim integrated routine)
  • LiquidCrystal
• Tooling: Arduino IDE
• Circuit design asset: Fritzing file in repository

Getting Started

Prerequisites

• Arduino IDE installed
• Board: Arduino Nano selected in IDE
• Sensor library available in IDE (see bundled library archive in repository)

Hardware Setup

1. Assemble components according to the circuit reference.
2. Use the BOM:
   • 1x Arduino Nano
   • 1x MAX30102
   • 1x LCD 16x2
   • 1x 10K potentiometer

Circuit references:

• Circuit image: assets/circuit.jpg
• Editable design: Fritzing Circuit.fzz

Circuit Design

Circuit design description:

• Arduino Nano reads MAX30102 via I2C (SDA/SCL)
• MAX30102 provides red/IR PPG signals for SpO2 and HR estimation
• 16x2 LCD is driven by digital pins for local telemetry
• 10K potentiometer is used for LCD contrast tuning

Firmware Deployment

git clone https://github.com/kershrita/Heart-Rate-and-SpO2-measurement.git
cd Heart-Rate-and-SpO2-measurement

1. Open code/code.ino in Arduino IDE.
2. Install required libraries if not already present.
3. Connect Arduino Nano via USB.
4. Compile and upload.

Runtime Flow

1. Place finger on sensor.
2. Wait for buffer fill and validity checks.
3. Read SPO2 and HR on LCD when stable; No Finger appears when contact is insufficient.

Results

Current repository demonstrates functional real-time measurement on-device with:

• Continuous PPG acquisition from red and IR channels
• On-board computation of SpO2 and heart-rate estimates
• Validity-gated LCD display outputs

Evaluation note:

• This repo focuses on embedded system integration and live inference.
• It does not currently include a labeled benchmark dataset, statistical error report, or clinical-grade validation protocol.

Model Details

• Model Type: deterministic physiological signal-processing algorithm (not a trained neural network)
• Selection Rationale: suitable for constrained hardware and real-time execution
• Inference Strategy: windowed estimation over synchronized red/IR signal buffers
• Quality Control: validity flags and finger-presence thresholding to suppress noisy states

License

Released under the MIT License. See LICENSE for details.