US 12,257,477 B2
Controller to determine swim characteristics of a swimmer and method thereof
Thorsten Sohnke, Asperg (DE); Amithash Kankanallu Jagadish, London (GB); Kaustubh Ashok Gandhi, Stuttgart (DE); Sengottuvelan Senthilmurugan, Erode (IN); Pudur Marudhachalam Dheepak, Coimbatore (IN); Mohan Pooja, Trivandrum (IN); and Tomar Mahima, Etah (IN)
Assigned to Robert Bosch GmbH, Stuttgart (DE); and Robert Bosch Engineering and Business Solutions, Bangalore (IN)
Appl. No. 17/905,079
Filed by Robert Bosch GmbH, Stuttgart (DE); and Robert Bosch Engineering and Business Solutions Private Limited, Bangalore (IN)
PCT Filed Feb. 8, 2021, PCT No. PCT/EP2021/052929
§ 371(c)(1), (2) Date Aug. 26, 2022,
PCT Pub. No. WO2021/170384, PCT Pub. Date Sep. 2, 2021.
Claims priority of application No. 202041008427 (IN), filed on Feb. 27, 2020.
Prior Publication US 2023/0142847 A1, May 11, 2023
Int. Cl. A63B 24/00 (2006.01)
CPC A63B 24/0006 (2013.01) [A63B 2220/40 (2013.01); A63B 2220/836 (2013.01); A63B 2244/20 (2013.01)] 7 Claims
OG exemplary drawing
 
1. A wearable device comprising:
at least one motion sensor including a multi-axis gyroscope or a multi-axis accelerometer; and
a controller operably connected to the at least one motion sensor and configured to determine swim characteristics of a swimmer, the controller configured to:
detect input signals from the at least one motion sensor,
perform stroke segmentations based on at least one of the detected input signals using a stroke segmentation module,
extract feature vectors through a feature extraction module from the detected input signals based on the stroke segmentations, and
determine the swim characteristics using the extracted feature vectors through a classifier module,
wherein the stroke segmentation module is configured to:
process at least one dominant axis signal from the detected input signals through a filter module and output a filtered signal;
generate an envelope signal from the filtered signal based on state machine conditions by:
following the filtered signal by default;
falling at a predefined rate when a value of the filtered signal is decreasing, referenced as a fall state, and
following the filtered signal when the value of the filtered signal increases and exceeds a value of the envelope signal in the fall state, referenced as a follow state, and
detect a stroke segment based on an occurrence of any one of two of the follow states and the fall states.