Wearable ECG Heart Monitor with Mobile Dashboard

Heart problems rarely announce themselves during a doctor's visit — they show up during the other 23 hours of the day, when nobody is recording. A wearable ECG monitor bridges that gap by continuously capturing the heart's electrical signal, and it is a superb electronics project because it forces the student through the whole analog-to-cloud chain: microvolt biopotential pickup, instrumentation amplification and filtering, ADC sampling, mains-notch filtering, and real-time wireless streaming to a dashboard. This project builds exactly that: an AD8232 single-lead analog front-end (1100× gain, 0.5–40 Hz band) amplifies the chest-electrode signal, an ESP32 samples it at 360 Hz, and a web dashboard renders the live PQRST trace with R-peak detection, heart-rate and HRV computed on the fly. The on-device OLED gives instant BPM feedback. The scope is stated plainly: an educational prototype for learning biomedical signal acquisition — not a diagnostic instrument.

1. Three gel electrodes on the adjustable chest band (RA, LA, RL) pick up the heart's microvolt electrical signal.
2. The AD8232 front-end amplifies it 1100× and band-limits it to 0.5–40 Hz, rejecting electrode offset and high-frequency noise.
3. The ESP32 samples the amplified output at 360 Hz with its 12-bit ADC and applies a 50 Hz mains notch plus baseline-wander removal in firmware.
4. A QRS detector finds R peaks; heart rate is computed from R-R intervals and HRV (SDNN) is accumulated over a rolling 60-second window.
5. Every 250 ms the ESP32 streams a packet (raw samples + BPM + signal-quality flag) over Wi-Fi to the web dashboard.
6. The dashboard renders the live trace with R-peak markers, the heart-rate trend, HRV and the event log; sessions export to CSV.
7. The on-device OLED mirrors the current BPM and a trace snippet for phone-free use.

Tech stack:

• AD8232 ECG analog front-end (1100× gain, datasheet)
• ESP32 devkit (12-bit ADC @ 360 Hz, Wi-Fi streaming)
• 3-lead chest electrodes + adjustable band
• 0.96" OLED display (on-device BPM)
• 1000 mAh Li-Po battery
• Pan-Tompkins-style QRS detection (firmware)
• Live web dashboard (HTML/CSS/JS, WebSocket)
• CSV session export

• [AD8232 analog front-end] Single-lead ECG amplification with 1100× gain and 0.5–40 Hz band filtering (datasheet values), designed for three chest electrodes (RA, LA, RL).
• [360 Hz sampling] The ESP32's 12-bit ADC samples the amplified signal at 360 Hz — the same rate as the classic MIT-BIH reference recordings — with a 50 Hz mains notch in firmware.
• [Live web dashboard] Streams the ECG trace in real time with R-peak markers, current BPM, HRV (SDNN over 60 s), session heart-rate trend and an event log.
• [On-device OLED] A 0.96" OLED on the band shows the live trace snippet and current BPM, so the device works without a phone in range.
• [R-peak detection] A Pan-Tompkins-style QRS detector runs in firmware; heart rate is computed from R-R intervals with a design-target accuracy of ±3 BPM.
• [Motion-artifact handling] Baseline wander is filtered and the dashboard flags noisy segments as poor signal quality instead of reporting fake heart rates.
• [Session export] Sessions can be exported as CSV (timestamp + ADC value) for offline analysis in the report.
• [Rechargeable wearable design] A 1000 mAh Li-Po powers about 14 hours of streaming (design target) in a comfortable adjustable chest band.

• Working wearable prototype (chest band, electrodes, AD8232, ESP32, OLED, battery)
• ESP32 firmware (sampling, filtering, QRS detection, Wi-Fi streaming)
• Live web dashboard with trace, analytics and CSV export
• Wiring diagram and assembly guide
• Project report PDF (biopotential background, signal-chain design, methodology, evaluation)
• PPT presentation for final review
• Viva Q&A preparation document (AD8232, sampling theory, QRS detection, HRV)

• This is an educational prototype, NOT a certified medical device — readings must never be used for diagnosis or treatment decisions. This is stated on the page, the dashboard and the report.
• ±3 BPM heart-rate accuracy is a design target, honestly stated — the report documents measured values after the prototype is tested.
• Single-lead ECG cannot show what a 12-lead clinical ECG shows; arrhythmia analysis is out of scope.
• Motion artifacts degrade the trace during exercise; the device is designed for rest or light activity.
• ~14 h battery life is a design target under continuous streaming; actual life depends on Wi-Fi conditions.