GymAware vs Accelerometer (PUSH)

Within-Unit Reliability and Between-Units Agreement of the Commercially Available Linear Position Transducer and Barbell-Mounted Inertial Sensor to Measure Movement Velocity – Mladen Jovanovic & Ivan Jukic (2020)


The purpose of this study was to investigate the within-unit reliability of GymAware linear position transducer (GYM) and PUSH2 inertial sensor to measure mean velocity (MV) and peak velocity (PV) during hexagonal barbell deadlift (HBD) and to examine the agreement between GYM and PUSH2 devices.

Twelve strength-trained men performed 2 HBD one-repetition maximum (1RM) sessions followed by 2 repetitions to failure assessments with 80 and 90% of daily 1RM. Barbell MV and PV were simultaneously monitored with 2 GYM and PUSH2 devices during all assessments.

An ordinary least products regression was used to assess within-units agreement and whether PUSH2 can accurately predict GYM velocity. In addition, residual standard error (RSE) and smallest detectable change in load (SDC%1RM) were also calculated.

GYM devices have been shown to be highly reproducible devices (RSE = 0.019–0.021 m·s−1; SDC%1RM = 1.795–2.679%). However, PUSH2 devices displayed a substantial amount of error (RSE = 0.133–0.220 m·s−1) and lack of sensitivity (SCD%1RM = 14.113–14.558%) to detect smallest change in load, which makes them untrustworthy for a regular use for monitoring athletes. Although very high correlations (r = 0.915–0.948) have been observed between PUSH2 and GYM velocity recordings, PUSH2 overestimated both MV and PV as indicated by high fixed and proportional bias.

The findings of the present study suggest that sport professionals should not use PUSH2 devices when the aim is to accurately monitor velocity variables during HBD exercise because low within-unit agreement and high fixed and proportional bias and RSE compared with GYM devices may compromise the utility of the collected data.

The Full Study