Learn · Glossary

Glossary

The same explanations that sit next to the figures in the tool — short and without unexplained jargon.

Value stream mapping

Lead time: The time a part takes from the first to the last process step — including all waiting and buffer time in between, not just the processing itself.
Value-added time: The sum of the pure cycle times of every process in the chain — excluding the waiting time sitting in inventory between them.
Process cycle efficiency: Value-added time divided by lead time, as a percentage. Shows how much of the total lead time is actually value-adding rather than waiting in inventory. The rating follows rule-of-thumb figures from the lean literature for discrete manufacturing — below roughly 5 % is the unimproved norm, continuous flow starts around 25 %. Guide values, not a standard.
Takt time: Available production time divided by customer demand. The pace at which one part must be finished for demand to be met — not a measured processing value, but a target.
Available production time: How many minutes per day are actually produced (e.g. one 8-hour shift = 480 min, two shifts = 960 min, less breaks). The other factor in the takt time formula alongside customer demand.
Actual output (exit rate): How many parts per day actually leave the value stream. Set by the slowest process (bottleneck / time trap) after its OEE losses: available production time ÷ bottleneck cycle time. Not to be confused with customer demand, which sets takt time — output is measured, takt is required. Together with WIP it gives the lead time: PLT = WIP ÷ output.
Balance chart (Yamazumi): One bar per station whose height shows its load including OEE losses, with takt time as a horizontal line. Unlike the timeline below the value stream, the axis here is the station rather than the sequence — which is what makes it readable which station breaks takt and where work content could be moved.
Capacity coverage: Actual output divided by customer demand. Below 100 % the line cannot meet demand: inventory then grows continuously, and lead time is no longer a stable figure but rises with every day.
Cycle time (C/T): The time a single process step takes to work on one part.
Changeover time (C/O): The time to switch a machine or process from one product or variant to the next.
OEE: Overall equipment effectiveness: the product of availability, performance and quality. Shows how much of the theoretically possible time is actually used in a value-adding way.
Operator count: How many people are needed to run this process step.
Pacemaker process: The one process in the chain that sets the pace for the whole production — usually the last one before shipping. Production control schedules only here; every upstream process replenishes by pull.
WIP (work in process): Inventory of unfinished parts between two process steps. Together with daily demand it determines how long a part waits at this point.
Value per piece: What one piece of inventory is worth — usually the production cost up to this point, not the sales price. This single factor is all that is missing to turn a count on the shelf into an amount of money.
Tied-up capital: Inventory times value per piece: the money sitting in the value stream instead of in the bank. Halve the inventory in a scenario and half of it comes back — that is the figure an investment rests on.
Inventory type: How inventory between two processes is controlled: standard (uncontrolled), supermarket (pull — replenished only on withdrawal), FIFO lane (fixed sequence, limited capacity) or continuous flow (no buffer, the part moves straight on).
Arrow type: How material flow is drawn: push (standard forwarding), pull (withdrawal from a supermarket), or shipment (delivery to/from supplier and customer). "Automatic" derives it from the inventory type.
Kanban type: Only relevant with a supermarket: a production kanban triggers replenishment at the supplying process, a withdrawal kanban only triggers the transport of an already produced part.
Supermarket: A controlled buffer stock from which the downstream process withdraws exactly what it needs via kanban — the upstream process only replaces what was withdrawn.
FIFO lane: A first-in-first-out connection between two processes with no free withdrawal — the processing sequence is strictly preserved, usually with limited capacity.
One-piece flow (continuous flow): Parts flow one at a time and without an intermediate buffer directly from one process to the next — no waiting, no inventory between the steps.
Bottleneck: A process whose effective cycle time (cycle time ÷ OEE) exceeds takt time — in its current state it cannot serve customer demand.
Heijunka (production levelling): Production mix and volume are spread evenly over time instead of running in large batches — it hangs off the pacemaker, because that is where production is actually scheduled.
Pitch (control time increment): A fixed time increment (a multiple of takt time) in which the pacemaker produces and releases a consistent quantity — it makes production control work in manageable portions rather than piece by piece.
Kaizen burst: A concrete improvement point marked at a process, needed to get from the current to the future state — not a general improvement wish, but a targeted note at exactly this spot.
Information flow: How production learns what to make and when: forecast and call-off from the customer, orders to the supplier, the schedule to the process. On the map in blue above the material flow, each arrow labelled with content and rhythm, e.g. “Weekly schedule · weekly”.
Electronic / manual: The two kinds of information flow. Electronic (system, e-mail, EDI) is drawn as a zigzag arrow, manual (paper, board, verbal) as a straight arrow.
Supermarket sizing: How many pieces a supermarket has to hold: cycle stock (replenishment interval × ADU) plus safety stock (covers variation) plus PLT stock (covers demand during production). Switching from push to pull replaces the stock that simply grew there with this calculated figure.
ADU (Average Daily Usage): The average daily usage of this part at this point, in pieces per day. Not the same as the project's annual demand, which describes what the customer takes at the end of the stream. Read the ADU from a usage extract over a representative period.
σ (standard deviation of ADU): How far daily usage swings around its mean, in pieces per day. It sets the safety stock. The calculator in the buffer panel works it out together with the ADU from a pasted series of daily usage figures.
PLT (process lead time): The process lead time in days, measured on the value stream. Sizing a supermarket counts the lead time without the share this buffer itself contributes — otherwise its own size would feed into its own precondition.

Capacity planning

Capacity analysis: The capacity traffic light for each line and month: does the available capacity cover demand? Uses the same figures as above on the process (cycle time, NEE — the same number as OEE), plus a shift model and monthly demand. Green is ok, orange/red means action needed, blue means underutilised.
Load rate: Demand ÷ available capacity of a line in one month. Below 0.5 underutilised, up to 1.0 ok, up to 1.2 watch, above that overloaded — the basis of the capacity traffic light.

Daily planning

PV — staff hours: All shift hours of a day according to the shift plan. Part-time counts with its hours, absences with 0. A night shift is split across both calendar days.
OT — order time: Staff hours spent on orders. Planned: staff hours × role weight (e.g. setters 1, apprentices 0.6). Actual: from time recording.
UT — machine hours: Machine run time. Planned from cycle time × order quantity, actual from operating hours of the host computer or confirmations.
Kmix — production mix coefficient: UT ÷ OT: machine hours per staff hour. Higher means the same crew looks after more machine time.
Kmix target: The level measured against: the average of the complete monthly Kmix of the previous year. Fixed from 1 January.
DLP — direct labour productivity: UT × f − OT, in hours. 0 means exactly at last year's level. Positive: the machine time carries more staff than scheduled. Negative: more staff than the machine time needs.
DLP factor f: 1 ÷ Kmix target. Converts machine hours into the staff hours they tie up at last year's level.
Realisation rate: Actual UT ÷ target UT from your own history. Makes the forecast more realistic than the plan target alone.
Complete month: A month with enough production days recorded (default 60 %). Only complete months count in the yearly average and in the following year's base.