The ‘sensing’ comes from data gathered by an array of sensors located on the device including multiple accelerometers, a magnetometer and a gyroscope. These sensors capture data that yield over 190 different parameters, such as orientation change, impact intensity and impact speed. If one of sensors notices an atypical pattern of activity, the pendant works to determine if an actual fall occurred through the following steps:
- Immediately after the sensors notice atypical activity, data is recorded for up to 10 seconds as needed.
- All of the recorded data is sent over the cellular network to the event classifier (located in a cloud-based server) for analysis.
- If the classifier determines that a fall has occurred, a call is automatically triggered from Bedford Medical Alert’s Care Team Responders.
The following graphs illustrate the difference between the data captured from a fall versus a drop.
- The first row is the accelerometer measured in g’s (1g‐ gravity force).
- The second row is the gyroscope (the rotation rate) measured in radiance per second.
- The third row is the magnetometer measured in microtesla.