IoT2023
SHEMS
A Smart Home Energy Management System built on Arduino and Raspberry Pi, monitoring and controlling residential energy consumption in real time through sensor firmware, a local aggregation layer, and a web dashboard. Built during my IoT Engineering internship at Intuz.
- Monitoring latency
- Real-time
- Hardware platforms integrated
- 2
- Hackathon recognition
- Winner
The problem
Household energy waste is invisible by design. The feedback loop is a bill that arrives a month after the behaviour that caused it, aggregated into a single number that tells you nothing about which device or which hour was responsible. You cannot change what you cannot see, and a monthly total shows you nothing you can act on.
The approach
Close the loop to seconds. Sensors on the circuits, firmware on an Arduino, aggregation on a Raspberry Pi sitting locally in the house, and a dashboard that shows consumption as it happens. Automated alerts fire on anomalous patterns, so the system surfaces the problem rather than waiting to be asked.
Energy waste in homes is invisible. Nobody knows they are leaving devices running until the bill arrives. I wanted to build a low-cost IoT system that makes residential energy use transparent and actionable in real time, not a month later.
Architecture
- 01
Sensing
Arduino firmware reading power draw from residential circuits.
- 02
Local aggregation
Raspberry Pi collecting and buffering sensor streams on-premises.
- 03
Analysis
Pattern detection over the aggregated series to identify inefficiency and anomalies.
- 04
Dashboard & alerts
Web dashboard for real-time visualisation, with automated alerts on anomalous draw.
Stack
- Hardware
- Arduino · Raspberry Pi
- Firmware & analysis
- Python · Pandas · Matplotlib
- Interface
- REST APIs · web dashboard
- Python
- Arduino
- Raspberry Pi
- IoT
- REST APIs
- Pandas
- Matplotlib
Outcome
Real-time visibility into residential energy use, with anomaly alerts that identify inefficiency patterns as they happen rather than a month later. The project was recognised as a hackathon winner.