SmartScan Node

Elder Care Occupancy Monitoring System using ESP32 and AWS IoT

Semi-Finalist — Northeastern University Husky Pitch Summit 2025

---

Architecture

ESP32 (HC-SR04 Sensor)
        |
        | MQTT over TLS/SSL (port 8883)
        v
AWS IoT Core
  - Thing: smartscan-node-aws-01
  - Topic: smartscan/presence
  - Rule: SELECT * FROM 'smartscan/presence'
        |
        v
AWS Lambda (presence_handler.py)
  - Parses occupancy payload
  - Derives data-driven insights on occupancy patterns
        |
        |---> DynamoDB (smartscan-occupancy)
        |       - device_id (PK), timestamp (SK)
        |       - distance_cm, occupied, TTL (30 days)
        |
        |---> SNS (smartscan-alerts)
                - Email alert on occupancy change
        |
CloudWatch
  - Lambda execution logs
  - IoT message throughput metrics

---

Project Structure

smartscan-node-aws/
├── firmware/
│   ├── main.py              # ESP32 MicroPython - sensor read + MQTT publish
│   └── config.example.py   # WiFi + AWS endpoint config template
├── lambda/
│   └── presence_handler.py # Lambda - DynamoDB write + SNS alert
├── terraform/
│   └── main.tf             # IaC - IoT Core, Lambda, DynamoDB, SNS, CloudWatch
└── docs/
    └── wiring.md           # ESP32 wiring guide

---

Hardware

Component	Model	Purpose
Microcontroller	ESP32-S3	WiFi, MQTT, sensor control
Presence sensor	HC-SR04	Ultrasonic distance (0-400cm)
LED	Onboard GPIO2	Visual presence indicator

Wiring:

• TRIG → GPIO 5
• ECHO → GPIO 18
• VCC → 5V
• GND → GND

---

AWS Setup

1. Provision infrastructure

cd terraform
cp terraform.tfvars.example terraform.tfvars
# Fill in alert_email
terraform init
terraform apply

2. Create IoT certificates

# In AWS Console → IoT Core → Security → Certificates
# Download: certificate.pem.crt, private.pem.key, AmazonRootCA1.pem
# Upload to ESP32 flash under /certs/

3. Flash firmware

# Install MicroPython on ESP32, then:
pip install ampy
cp firmware/config.example.py firmware/config.py
# Edit config.py with your WiFi + AWS endpoint
ampy --port /dev/ttyUSB0 put firmware/main.py
ampy --port /dev/ttyUSB0 put firmware/config.py
ampy --port /dev/ttyUSB0 mkdir /certs
ampy --port /dev/ttyUSB0 put certs/AmazonRootCA1.pem /certs/AmazonRootCA1.pem
ampy --port /dev/ttyUSB0 put certs/certificate.pem.crt /certs/certificate.pem.crt
ampy --port /dev/ttyUSB0 put certs/private.pem.key /certs/private.pem.key

---

Data Flow

1. ESP32 reads HC-SR04 every 5 seconds
2. Debounce logic (3 consecutive reads) prevents false triggers
3. Presence payload published to smartscan/presence via MQTT over TLS
4. IoT Core rule triggers Lambda on every message
5. Lambda writes to DynamoDB with 30-day TTL
6. SNS alert fires on occupancy state change
7. CloudWatch logs Lambda execution and IoT message throughput

---

Payload Schema

{
  "device_id":   "smartscan-node-aws-01",
  "distance_cm": 87.3,
  "occupied":    true,
  "timestamp":   1714204800
}

---

Motivation

Designed for elder care facilities where passive, non-invasive room occupancy monitoring can improve staff response times and resident safety. The system runs entirely on edge hardware with a serverless cloud backend — no continuous cloud polling, minimal cost.

---

Tech Stack

Layer	Technology
Edge	ESP32-S3, MicroPython, FreeRTOS
Connectivity	MQTT over TLS/SSL (port 8883)
Cloud ingestion	AWS IoT Core
Processing	AWS Lambda (Python 3.12)
Storage	AWS DynamoDB (PAY_PER_REQUEST)
Alerting	AWS SNS (email)
Monitoring	AWS CloudWatch
IaC	Terraform

---

License

MIT