Model a value stream once and get current-state metrics, a future-state design and printable reports that all agree with each other. The editor runs in this page and needs JavaScript; the reference it links to does not.
In most mapping tools the drawing lives in one place and the numbers in another, and nothing guarantees that the figure in the presentation is the figure on the screen. Here the on-screen metrics, the A3 print, the PDF reports, the workbook and the image export are produced by one calculation engine. Change an input and all of them change together.
Every input carries whether it was observed or estimated, and findings are reported at lower confidence on a map built mostly from estimates. Where all three factors have not been measured the product will not call the result OEE — it reports an A×Q score and names the component it is missing. The benchmark includes a case with no answer in it, because declining to answer is part of what is being verified.
The engine reproduces three published totals from Rother & Shook's Learning to See — takt 60 s, processing time 188 s, production lead time 23.6 days — and reaches the example's central finding, that a 62-second assembly cannot meet a 60-second takt. What that does and does not prove is set out in full on the benchmark page.
The interface, the findings and every exported report are written in both languages, and the PDF prints Turkish characters correctly. Each of the thirteen figures the product calculates is defined with its formula in the glossary.