Fewer false alarms. Smarter home security.
SmartNest Secure links low-cost IoT hardware with a deep learning model that recognises whether a person on camera is a resident or an intruder from their walking pattern.
Home alarms that cry wolf
Most home systems detect motion, but they cannot tell a resident from a stranger.
False notifications
Owners get alerts every time someone at home walks past a sensor.
Separate gadgets
Appliance control, climate monitoring and security often need different systems.
Little intelligence
Basic surveillance records events but does not decide what matters.
One kit, three jobs, one smart model
SmartNest Secure combines remote control, environment sensing and intelligent intruder detection.
Gait-based intruder detection
- Classifies walking patterns
- Resident or intruder
- Cuts false alarms
Remote appliance switching
- ESP8266 controller
- Control over the internet
- Low-cost parts
Indoor monitoring
- DHT11 sensor
- Response time modelled
- Supports weather prediction
How SmartNest Secure works
From a movement in the house to the right response.
- 1
Detect motion
A PIR sensor notices movement in or around the home.
PIRTrigger - 2
Capture
The ESP32 camera records the person.
ESP32-CAM - 3
Classify
A CNN reads the walking pattern and labels the person.
CNNDeep learning - 4
Act
Owners are alerted only for intruders and can control appliances remotely.
RelayIoT
What the research showed
Results from the prototype described in the published study.
| Component | Role |
|---|---|
| ESP32 camera | Surveillance video |
| PIR motion sensor | Movement detection |
| ESP8266 board | Control and connectivity |
| 5 V four-channel relay | Appliance switching |
| DHT11 sensor | Temperature and humidity |
Source: Taiwo et al., Wireless Communications and Mobile Computing 2022(1), 9307961 (2022). doi:10.1155/2022/9307961
Who SmartNest Secure is for
Built for homes and small buildings that want security without the noise.
Homeowners
- Fewer false alarms
- Control appliances remotely
- Watch indoor comfort
Landlords and facility managers
- Low-cost hardware
- Monitor several units
- Add-on to existing wiring
Integrators
- Open, common components
- Model can be retrained
- Extend with new sensors
Where the evidence stands
SmartNest Secure exists as a working research prototype.
Prototype stage
The system was built and tested as a lab prototype.
Gait data
Accuracy depends on learning each household's walking patterns.
Privacy design ahead
Camera data handling and consent need careful product design.
The team
Founded in 2022 by the research team behind the study.
Mubarak S. Almutairi
Co-author of the SmartNest Secure research.
Olutosin Taiwo
Co-author of the SmartNest Secure research.
Absalom E. Ezugwu
Co-author of the SmartNest Secure research.
Olaide N. Oyelade
Co-author of the SmartNest Secure research.
Frequently asked questions
How does it tell residents from intruders?
A CNN model classifies people by their walking pattern.
How accurate is it?
The classification model reached 99.8% accuracy in the study.
What hardware does it use?
An ESP32 camera, PIR sensor, ESP8266 board, 4-channel relay and DHT11 sensor.
Who is behind it?
Researchers including Mubarak S. Almutairi.
Peer-reviewed science behind SmartNest Secure.
Read the full study behind the product.
Taiwo, O., Ezugwu, A. E., Oyelade, O. N., & Almutairi, M. S. (2022). Enhanced Intelligent Smart Home Control and Security System Based on Deep Learning Model. Wireless Communications and Mobile Computing, 2022(1), 9307961.
https://doi.org/10.1155/2022/9307961