How to make a smart doorbell via AI

The system uses the Renesas RA4M1-based Arduino UNO R4 WiFi (€25,00) and a MAX4466 Electret Microphone Amplifier ($6.95), to be precise, to add connectivity to your bell. It will detect when the doorbell is ringing and could then send you a message, for example.

Smart doorbell

The creator – mad_mcu – takes you through the process, using the free tool NanoEdge AI Studio to create an AI model able to classify if the doorbell is ringing or if it is just background noise.

Having connected the microphone to the Arduino board, and with a USB data cable connecting the board to a PC, you can start to collect microphone data.

The goal is to use NanoEdge to create an “Anomaly detection project”, detecting the doorbell as “normal” and everything else as an “abnormal” behavior.

Then, once you have the anomaly detection library setup, you need to add it to the Arduino code…

You can read more on the Arduino Project Hub, including code for the data logger and the main smart doorbell system.

See also: Arduino PLC Starter Kit plugs into industrial automation