PNSQC 2013 - Kathy Iberle
summary
Argues that applying queuing theory to test labs — counting batches, drawing a cumulative flow diagram, and using a visual board — finds hidden wait time and lets teams cut work-in-process limits and shrink batches. Shows concrete wins: splitting tests into 1–2 hour batches to avoid night/day handoffs, moving people from new features into fixes to create a visible hockey‑stick improvement, and quarterly cadence that raises improvement throughput by roughly 50%.
AI-written summary · about
chapters (25)
- 0:00Intro
- 1:12INTRODUCTION
- 2:28LEAN: THREE BODIES OF KNOWLEDGE
- 3:05THE SYSTEMS VIEW
- 5:15END OF DAY 1
- 5:32MIDDLE OF DAY 2
- 6:01START OF DAY 3
- 7:05AN AVERAGE DAY
- 8:44MATH TO THE RESCUE!
- 9:38WHAT ERLANG (AND OTHERS) LEARNED It is possible to do more with the same capacity.
- 14:18LEAN AND QUEUEING THEORY
- 21:36WHAT IS A BATCH?
- 23:24WHAT VALUE DOES A TEST LAB PRODUCE?
- 24:09TYPICAL BATCHES IN A SOFTWARE TEST LAB
- 27:25A BATCH HAS A START AND AND AN END
- 30:24MAKE YOUR ACTIVITIES VISIBLE
- 31:43VISIBILITY: A VISUAL PLANNING BOARD
- 35:40TRACK PROGRESS OVER TIME
- 36:58WHAT YOU CAN SEE ON THE CFD
- 39:45EXTRAPOLATION OF PROGRESS
- 48:01A PREDICTABLE SYSTEM CAN BE OPTIMIZED
- 50:08VALUE-STREAM MAPPING
- 51:14OPTIMIZE A VARIABLE SYSTEM Monitor and adjust in real-time
- 53:10SOFTWARE TEST - A BATCH
- 57:36SMALL BATCHES → PREDICTABILITY
