# US 2025/0375660 A1 · Exercise machine struggle detection > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2025/0375660 A1 - Document ID: US20250375660A1 - Status: Published application - Kind: A1 - Application number: 19/232,700 - Filed: 2025-06-09 - Published: 2025-12-11 - Earliest claimed: 2022-01-17 - Assignees: Tonal Systems, Inc. - Applicants: Tonal Systems, Inc. - Inventors: Brandt Belson; Pavel Katkov; Egor Ponomarev; Giuseppe Barbalinardo - Web page: https://patentgrep.com/patent/US20250375660A1 ## CPC - A63B 24/0006: {Computerised comparison for qualitative assessment of motion sequences or the course of a movement} - A63B 24/0062: {Monitoring athletic performances, e.g. for determining the work of a user on an exercise apparatus, the completed jogging or cycling distance} - G06F 17/18: for evaluating statistical data {, e.g. average values, frequency distributions, probability functions, regression analysis (forecasting specially adapted for a specific administrative, business or logistic context G06Q10/04)} - G06N 3/044: Recurrent networks, e.g. Hopfield networks - G06N 3/045: Combinations of networks - G06N 3/08: Learning methods - G06N 3/082: modifying the architecture, e.g. adding, deleting or silencing nodes or connections - G06N 5/01: Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound - A63B 2024/0009: {Computerised real time comparison with previous movements or motion sequences of the user} - A63B 2024/0068: {Comparison to target or threshold, previous performance or not real time comparison to other individuals} - A63B 2220/51: Force - A63B 2220/833: Sensors arranged on the exercise apparatus or sports implement - G06N 20/20: Ensemble learning - A63B: APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT (apparatus for passive exercising, massage A61H) - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract Performance information associated with a previous repetition of an exercise movement is received. Performance of one or more upcoming repetitions is predicted, based at least in part on the performance information associated with the previous repetition of the exercise movement. A failure classification of whether the one or more upcoming repetitions is associated with an occurrence of physical failure is performed, based at least in part on the predicted performance of the one or more upcoming repetitions of the exercise movement. A number of repetitions in reserve is determined, based at least in part on the failure classification. ## Claims ### Claim 1 A method, comprising: receiving performance information associated with a previous repetition of an exercise movement for a user; predicting struggle performance of one or more upcoming repetitions for the user based at least in part on the performance information associated with the previous repetition of the exercise movement and a group historical workout set for a plurality of users; 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 struggle performance of the one or more upcoming repetitions of the exercise movement and a user historical workout set for the user; and determining a number of repetitions in reserve based at least in part on the failure classification. ### Claim 2 (depends on 1) The method of claim 1, wherein the one or more upcoming repetitions is between 1 and 5 upcoming repetitions. ### Claim 3 (depends on 1) The method of claim 1, wherein the performance information comprises at least one of the following: maximum concentric velocity of the previous repetition, maximum eccentric velocity of the previous repetition, change in maximum concentric/eccentric velocity between the previous repetition and a preceding repetition, change in maximum concentric/eccentric velocity from a set average, average concentric/eccentric velocity of the previous repetition, change in average concentric/eccentric velocity from the previous repetition, change in average concentric/eccentric velocity from the set average, duration of concentric/eccentric phase of the previous repetition, change in duration from the previous repetition, change in duration from the set average, range of motion for the previous repetition, change in range of motion from the previous repetition, change in range of motion from the set average, mid-repetition pause, mid-repetition pause in previous rest, mid-repetition pause in average repetition, distance covered, user features, relative muscle volume, rest time before repetitions, repetition goal, elapsed duration, movement information, user information, exercise movement form information, and spotter engagement. ### Claim 4 (depends on 1) The method of claim 1, wherein predicting struggle performance comprises using at least one of the following: a deep learning model architecture, a lasso regression, and a linear regression. ### Claim 5 (depends on 1) The method of claim 1, wherein predicting struggle performance comprises using a sequence model architecture. ### Claim 6 (depends on 1) The method of claim 1, wherein predicting struggle performance comprises using at least one of the following: a recurrent neural network (RNN) architecture, and a diluted neural network architecture. ### Claim 7 (depends on 1) The method of claim 1, wherein predicting struggle performance comprises using at least one of the following: a long short-term memory (LSTM) architecture, and a WaveNet architecture with a reduced number of hidden layers. ### Claim 8 (depends on 1) The method of claim 1, wherein predicting struggle performance comprises using a natural language upcoming word prediction architecture. ### Claim 9 (depends on 1) The method of claim 1, wherein predicting struggle performance comprises using a small neutral network architecture contained inside a mobile tablet environment. ### Claim 10 (depends on 1) The method of claim 1, wherein predicting struggle performance comprises using a training set based at least in part on another repetition. ### Claim 11 (depends on 1) The method of claim 1, wherein predicting struggle performance comprises using a filtered training set based at least in part on another repetition. ### Claim 12 (depends on 1) The method of claim 1, wherein predicting struggle performance comprises an output with at least one of the following: a range of motion for the one or more upcoming repetitions, a maximum concentric/eccentric velocity for the one or more upcoming repetitions, an average concentric/eccentric velocity for the one or more upcoming repetitions, and a duration of a concentric/eccentric phase for the one or more upcoming repetitions. ### Claim 13 (depends on 1) The method of claim 1, wherein performing a failure classification comprises using at least one of the following: a heuristic, a logistic regression architecture, and a random decision forest architecture. ### Claim 14 (depends on 1) The method of claim 1, wherein performing a failure classification comprises an output with at least one of the following: a label of a successful repetition, and a label of a physically failed repetition. ### Claim 15 (depends on 1) The method of claim 1, wherein performing a failure classification comprises using symptoms of at least one of the following: reduced velocity during a concentric phase, increased velocity during an eccentric phase, longer rests between repetitions, reduced range of motion. ### Claim 16 (depends on 1) The method of claim 1, wherein performing a failure classification comprises using prior information of at least one of the following: user information regarding sleep, relative muscle volume for a currently used muscle group, and exercise movement form information. ### Claim 17 A system, comprising: a processor configured to: receive performance information associated with a previous repetition of an exercise movement for a user; predict struggle performance of one or more upcoming repetitions for the user based at least in part on the performance information associated with the previous repetition of the exercise movement and a group historical workout set for a plurality of users; perform 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 struggle performance of the one or more upcoming repetitions of the exercise movement and a user historical workout set for the user; and determine a number of repetitions in reserve based at least in part on the failure classification; and a memory coupled to the processor and configured to provide the processor with instructions. ### Claim 18 (depends on 17) The system of claim 17, wherein the one or more upcoming repetitions is between 1 and 5 upcoming repetitions. ### Claim 19 A computer program product embodied in a non-transitory computer readable medium and comprising computer instructions for: receiving performance information associated with a previous repetition of an exercise movement for a user; predicting struggle performance of one or more upcoming repetitions based at least in part on the performance information associated with the previous repetition of the exercise movement and a group historical workout set for a plurality of users; 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 struggle performance of the one or more upcoming repetitions of the exercise movement and a user historical workout set for the user; and determining a number of repetitions in reserve based at least in part on the failure classification. ### Claim 20 (depends on 19) The computer program product of claim 19, wherein the one or more upcoming repetitions is between 1 and 5 upcoming repetitions.