Sign in Subscribe
Machine learning

In-Sensor Fall Detection Patent Application Filed by STMicroelectronics International N.V.

Strategic Insights

— 1 min read

STMicroelectronics International N.V., a leading semiconductor manufacturer, has filed a patent application for a novel in-sensor fall detection system. According to the application, the system is designed to detect falls using a pressure sensor and an inertial measurement unit (IMU), eliminating the need for machine learning techniques that consume high power. The proposed solution is ideal for portable wearable devices, such as headphones, smart watches, belts, and necklaces.

Key Takeaways:

  • The in-sensor fall detection system is designed to execute in the sensor hardware itself, reducing power consumption and enabling continuous execution within portable wearable devices.
  • The system utilizes a finite state machine to perform shock detection, altitude change detection, shock end detection, and steadiness detection, eliminating the need for machine learning techniques.
  • The proposed solution is comprised of a pressure sensor, an IMU, and an application processor, where the IMU is communicatively coupled between the pressure sensor and the application processor.
  • The IMU generates acceleration measurements, detects falls using both pressure sensor data and accelerometer data, and outputs an indication of the fall event.
  • The system claims that processing of the application processor consumes a greater amount of power than processing of the IMU, making it an efficient solution for portable wearable devices.
  • The proposed method of claim 10 comprises generating pressure measurements, generating acceleration measurements, detecting a fall event based on the pressure measurements and the acceleration measurements, and outputting an indication of the fall event.
  • The system detects a shock state, an altitude change, an end of the shock state, and a steady state, and the detecting of the fall event is based on the shock state, the altitude change, the end of the shock state, and the steady state.

Statistics:

  • The patent application was filed on April 28, 2025.
  • The patent covers a method for in-sensor fall detection using a pressure sensor and an IMU.
  • The proposed system is designed to reduce power consumption and enable continuous execution within portable wearable devices.
  • The system utilizes a finite state machine to perform shock detection, altitude change detection, shock end detection, and steadiness detection.
  • Processing of the application processor consumes a greater amount of power than processing of the IMU.

Sources:

  • BRACCO, Lorenzo; RIZZARDINI, Federico. In-Sensor Fall Detection. U.S. Patent Application Number 20250252837, filed April 28, 2025 and posted August 7, 2025. Patent URL (for desktop use only): https://ppubs.uspto.gov/pubwebapp/external.html?q=(20250252837)&db=US-PGPUB&type=ids

Read more