Search AI Products and News

The safety risks associated with lithium battery fires are often a cause for concern. To address this, scientists have proposed a method to provide early warnings of battery fires using sound. The study discovered that lithium-ion batteries undergo a series of chemical reactions before catching fire, leading to a gradual increase in internal pressure and ultimately causing the battery to swell. Due to the hard casing of the batteries, which cannot accommodate this swelling, the safety valve inside the battery bursts when the pressure becomes too high, creating a distinct sound.

Google's HeAR bioacoustic model detects various diseases such as tuberculosis, COVID-19, and chronic obstructive pulmonary disease by analyzing sounds, trained on 313 million audio clips, demonstrating exceptional diagnostic capabilities. HeAR outperforms existing models in multiple health acoustic events, cough, and lung function inference tasks, particularly excelling in detecting COVID-19 and tuberculosis. Salcit Technologies has applied HeAR in its product Swaasa® for lung assessment.