Bar speedRep 1
0.00m/s
Setting up
Best rep—
Loss—
Est. RIR—
Load · back squat
Stop the set at
20% velocity loss
Speed sensor
Mean concentric velocity per repMeasured
Load-velocity profile · todayEstimated max
This set
Depth per rep · vs parallel
Front view
Six weeks · illustrative lifter

The trend matters more than any one number

Estimated 1RM and weekly volume loadEstimated

e1RM up about 6% in six weeks, with a deload in week 4

Each point is that week's load-velocity estimate, with the band from the minimum-velocity uncertainty. Volume load is sets × reps × kg, counted automatically from the rep log. A true heavy single, when one happens, recalibrates the line.

How the method works

Speed tells you how heavy it really was

  1. Bar speed falls as the load rises, almost in a straight line (R² 0.96 in 80 trained men, Sánchez-Medina 2017).
  2. Five warm-up sets give this lifter's own line. Extending it to the speed of a true max, about 0.32 m/s for the squat (range 0.24–0.37), gives an estimated 1RM.
  3. Within a set, each rep slows as fatigue builds. Stopping at a set velocity loss caps fatigue: 20% gave similar strength to 40% with 40% fewer reps (Pareja-Blanco 2017).
  4. How fast the stop rep is, relative to the lifter's own profile, estimates reps in reserve. It is worked out once per set, shown ±2, and only trusted after 2–3 calibration sessions.
  5. Linear extrapolation tends to read high in free-weight squats. Room shows a range and favours the trend.
Camera only · real footageReal tracking

The same read on a real lifter

0
—
—

A side-on barbell row from one ordinary camera. The line is the bar's path, traced from the hands; the white tick is mid-foot. Reps count from the bar's height, and the torso angle shows whether the lifter stands up to cheat the weight.

Licensed stock clip with pose tracking at 15 fps. No load tag here, so no speed or estimated max: bar drift is in body heights, not centimetres.

Team rack mode0 reps loggedSimulated

Eight racks, one coach, one live board

#AthleteBest repm/sReadiness

Simulated session: back squat at the prescribed 75%. The board sorts by each athlete's best rep speed this set. Readiness is today's 60 kg warm-up speed against that athlete's own four-week average.

  • Roster auto-assign. The team roster loads before the session; each athlete taps in at a rack and the camera holds them for the session.
  • Readiness from the warm-up. More than about 6% slower than their own average on the warm-up bar, and the coach sees "watch" or "drop 5–10%" before the working sets start.
  • Live leaderboard. Bar speed per rep on the wall for all eight racks, so velocity-loss stops happen without a coach at every rack.
  • Export. Sets, reps, load and speed go to the team's program after the session.
TeamBuildrBridge AthleticCSV
Why this and not a rack-only system: the same kit runs the class floor, the testing-day combine and the athlete's verified profile. One install, three jobs.
Sensor honesty

Which numbers you can trust, by sensor

SensorBar speedGood for
Camera onlyApproximatephone-camera apps read 0.06–0.13 m/s offReps, tempo, depth, bar path, trends
Depth camera per rackGoodbar path, depth and tempo near-certainDaily VBT, velocity loss
Bar sensor or tetherVerifiedlinear transducer class, about 5% CVe1RM testing, research-grade speed