NAGARE automates cycle time measurement at manual stations, timing every cycle from video instead of a short stopwatch sample. Each start and finish is detected the same way on every shift, so averages, spread and the slowest cycles are visible for each station and operation.
Manual cycle times vary with the operator, the variant and the hour. A study of twenty cycles captures one moment, and the line is then balanced on numbers that no longer match reality.
Our AI process monitoring software, NAGARE™, performs cycle time measurement continuously by detecting when each cycle and operation starts and ends on camera. Results are timed consistently across operators and shifts, with no one standing at the station. The productivity figure is a NAGARE product-level metric. Confirm any effect on your measured station during the trial.
Cycle time measurement with NAGARE relies on a consistent definition of where each cycle begins and ends, applied by a model to every cycle the camera sees.
Engineering agrees which action begins and ends a cycle and each operation, using recorded cycles from the station.
The model detects those actions on every cycle and records the time for the cycle and its operations.
Results show median, spread and slow cycles by station, variant and shift, with clips for the outliers.
Send a video of one station and your current cycle figures. We will show how measured cycles compare with the study.
Get In TouchEach method produces a number. They differ in how many cycles it covers, how consistent it is and whether you can see why a cycle ran long.
At APA Engineering, NAGARE monitors cycle time for every step and every kit across its kitting stations, alongside real-time sequence and quantity verification.
APA Engineering's published results cover its complete export kitting operation, of which cycle monitoring is one part.
Cycle time is how long a station or operator actually takes to complete one unit. Takt is the pace required to meet customer demand, set by available working time divided by demand. A station whose cycle exceeds the required pace becomes a constraint on the whole line. Comparing the two, station by station, is the starting point for balancing work and deciding where improvement effort should go first.
A manual timing study captures a limited number of cycles, often with the operator aware of being watched. It is useful for setting up a method, but it misses rare long cycles, changes across shifts and drift after the study. Continuous cycle time measurement covers every cycle, so the figures reflect normal work rather than a few observed minutes. Studies remain valuable when designing a new method.
The start and end points must be the same every time, or results become noise. A common choice is to time from the start of one action to the start of the same action on the next unit, because operators often reorder the steps in between. Each work element inside the cycle can be timed in the same way, so long cycles can be traced to the element that caused them.
Two stations with the same average can behave very differently. One may be steady while the other alternates between quick cycles and long stalls that starve the next station. Cycle-to-cycle variation, the median cycle and the standard deviation together show which stations are predictable and which need attention, and they make an improvement easier to prove afterwards.
Balancing the line needs a consistent definition of the work and comparable product mixes. Measured cycles can highlight a candidate bottleneck station, but confirm it against waiting, buffers, machine availability and the surrounding line before moving tasks. Compare repeated cycles and relevant shifts, then review whether the change improves output without moving the constraint elsewhere. Video and balance charts support that investigation. A long average alone does not prove which station limits the whole line.
Machines may report cycles through their control system while manual stations report little. Camera-derived observations can help close that gap. Combining them with MES timestamps requires agreed interfaces, synchronized clocks, unit or batch context and comparable station cycle time definitions. Confirm which connections and views are available in the proposed deployment rather than assuming machine data joins automatically. A shared timeline is useful only when the event meanings and any missing observations are understood by the team.
Linking each long cycle to its video clip shows the reason behind it: a missing part, a difficult fit, a tool out of reach or a method that differs between operators. That turns a timing report into a list of improvement ideas. Supervisors review a handful of clips instead of standing at the station, and engineers can test a fix and see its effect quickly. Over a few weeks, the clips also show which causes are rare and which repeat on every shift.
Send a video of the station across a normal shift, the current takt and any existing study figures, and the variants built there. Tell us which stations you think limit output. With that, we can agree the cycle start and end definition, propose camera positions and show the first measured results for review.
Share a station video, your takt and the current study figures. We will plan cycle time measurement for that line and explain timing and cost. Most measurement projects are producing data within 4 weeks.
