We have implemented a smart baby monitor that records the temperature of a baby regularly and saves it on an online database, Firebase. It uses machine learning to check if the baby is crying or not.

An android user-friendly mobile application has been made so that the parents can receive notifications and check on the baby’s current temperature and its history using Java as a programming language.

The ARM microcontroller is programmed to collect the read the data from the sensor using I2C protocol and C++ as a programming language. Furthermore, the audio recording is continuously processed every 5 seconds to determine if the baby is crying or not using a CNN and uTensor.

The monitor sends the collected data to an online database, Firebase, via HTTP request.

The following video presents the functionalities of the mobile application.

• machine learning
• mobile application
• real-time database
• embedded electronics