Hablo is an IoT-based smart system designed to interpret baby cries and translate them into meaningful insights about a child’s needs. The project combines embedded hardware, signal processing, and machine learning to deliver real-time analysis of audio signals and provide actionable feedback to parents through a connected application.

In this project, I was responsible for designing and building the electronic hardware and developing the associated firmware. This included creating a system capable of accurately capturing, sampling, and processing audio signals in real time. I worked closely with a dedicated software and design team to bring the full product to life, ensuring seamless integration between hardware, cloud infrastructure, and user-facing applications.

The system performs initial signal conditioning and feature extraction at the edge, enabling low-latency response. A convolutional neural network (CNN), developed using TensorFlow, analyzes the harmonic structure of baby cries to classify them into different states such as hunger, discomfort, or fatigue.

The device connects to a mobile application via the internet, allowing parents to monitor and configure the system remotely. Communication between the embedded system and the cloud is handled through efficient socket-based protocols.

From concept to prototype, the project covered the full development cycle: schematic design (KiCad), PCB prototyping, 3D-printed enclosure design, firmware development, and system validation. Python was used for data processing, model development, and integration, deployed on a Raspberry Pi platform.

Developed within a multidisciplinary team, Hablo demonstrates the integration of embedded systems, AI, and IoT into a functional consumer product.

Technologies: KiCad, Python, Raspberry Pi, 3D Printing, TensorFlow, Embedded AI, IoT

Recognition: 2nd Place — Orange Summer Challenge 2018