SENSA

According to the World Health Organisation (2023), 466 million people globally have disabling hearing loss — and the number is rising. By 2050, that figure is projected to exceed 900 million, nearly 1 in 10 people on the planet.

For deaf and hard-of-hearing individuals, missing a smoke alarm, a crying baby, or a door knock isn't an inconvenience — it's a safety risk and an independence problem.

Sensa runs a continuous, low-latency audio analysis loop on your device. It captures ambient sound via the microphone, processes amplitude patterns in a 3-second ring buffer, and matches against 22 known sound signatures — all locally, in real time, without ever leaving your phone.

When a match is detected, Sensa triggers an immediate vibration alert. A background service keeps monitoring even when the screen is off.

Sensa ships with a comprehensive library of environmental sounds mapped into four distinct categories. Each sound uses a unique amplitude-pattern signature tuned for accuracy in real home environments.

Sensa's detection engine doesn't use complex neural networks or cloud inference. Instead, it uses a lightweight, battery-efficient temporal amplitude pattern analysis pipeline — purpose-built for mobile devices running 24/7 in the background.

Each sound signature is described as an amplitude envelope over time. The detection loop samples the microphone at 50ms intervals, maintains a 3-second sliding window of amplitude values, and runs pattern matching every 300ms against all active sound signatures.

Commercial sound-alert products exist — but they require separate hardware, recurring subscriptions, and often cloud connectivity. Sensa delivers comparable awareness using the smartphone already in your hand.

Sensa processes only a single numeric amplitude value per 50ms tick — never actual audio content. No recording, no file storage of audio, no server. The microphone is used exclusively for real-time analysis.