US 12,357,876 B2
Exercise machine struggle detection
Brandt Belson, San Francisco, CA (US); Pavel Katkov, Samara (RU); Egor Ponomarev, Samara (RU); and Giuseppe Barbalinardo, Berkeley, CA (US)
Assigned to Tonal Systems, Inc., San Francisco, CA (US)
Filed by Tonal Systems, Inc., San Francisco, CA (US)
Filed on May 31, 2023, as Appl. No. 18/204,083.
Application 18/204,083 is a continuation of application No. 17/714,045, filed on Apr. 5, 2022, granted, now 11,701,546.
Claims priority of provisional application 63/300,235, filed on Jan. 17, 2022.
Prior Publication US 2023/0414997 A1, Dec. 28, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. A63B 24/00 (2006.01); G06F 17/18 (2006.01); G06N 3/045 (2023.01)
CPC A63B 24/0006 (2013.01) [A63B 24/0062 (2013.01); G06F 17/18 (2013.01); G06N 3/045 (2023.01); A63B 2024/0009 (2013.01); A63B 2024/0068 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method, comprising:
creating a neural network architecture;
training the neural network architecture at least in part with performance information associated with a previous repetition of an exercise movement;
predicting performance, including a predicted speed, of one or more upcoming repetitions based at least in part on the performance information, including a previous speed, associated with the previous repetition of the exercise movement by labeling the one or more upcoming repetitions using the neural network architecture based at least in part on features of a current repetition;
wherein the previous speed is measured using a sensor coupled to an actuator for the exercise movement, and wherein the sensor is configured to measure position of the actuator;
performing a failure classification of whether the one or more upcoming repetitions is associated with an occurrence of physical failure based at least in part on the predicted speed of the one or more upcoming repetitions of the exercise movement;
determining a number of repetitions in reserve based at least in part on the failure classification;
determining a future suggested weight based at least in part on the number of repetitions in reserve; and
controlling a motor controller to adjust a perceived weight stack associated with a motor to the future suggested weight.