Conducting an OEE loss-tree analysis
An OEE loss-tree analysis breaks overall equipment effectiveness down into its underlying availability, performance and quality losses, then decomposes each into the specific loss categories driving it. It turns a single headline OEE number into a ranked, actionable map of where production capacity is actually being lost.
What it is
OEE combines availability, performance and quality into one figure, but the figure alone does not tell you what to fix. A loss tree structures the analysis: it splits the gap from perfect OEE into the major loss buckets — breakdowns, changeovers, minor stops, speed loss, defects and start-up rejects — and quantifies each, so improvement effort can be aimed at the largest contributors rather than guessed at.
Why it is done
Tracking OEE without analysing its losses produces a number that goes up and down with no clear cause and no obvious action. The loss tree converts measurement into direction by showing which losses dominate, revealing that, for example, minor stops or speed loss may be quietly costing more than the visible breakdowns. It is what makes OEE a tool for improvement rather than just a score.
How it is done
Production and downtime data are collected at sufficient detail to attribute time to specific loss categories, which often means improving how stops and slow running are recorded in the first place. The losses are then quantified and arranged into the availability, performance and quality branches and ranked by magnitude. The largest losses become the targets for focused improvement work, and the tree is refreshed as data improves and losses shift.
- Collect detailed loss data
- Categorise each loss
- Build the loss tree
- Quantify each branch
- Rank by magnitude
- Target the largest
What to watch for
Coarse or inconsistent loss recording makes the tree misleading, especially under-recording minor stops and speed loss that are easy to overlook. Chasing the most visible loss rather than the largest measured one is the common misdirection a proper loss tree is meant to prevent.
Related practices
Reducing changeover time with SMED
Optimising clean-in-place (CIP)
Running a process capability study
Related topics
OEE (Overall Equipment Effectiveness) · Unplanned Downtime · Bottleneck
Common in: Food Processing · Brewing & Beverage · Dairy · Paper & Packaging · Pharmaceuticals