Process Flow Analysis

Maps. describes, and identifies potential improvement opportunities

  1. Identify start/end points

  2. Identify decision points, buffers, and activities

  3. Collect data on process at each identified element

  4. Analyze using Little’s Law or similar

  5. Calculate performance metrics

  6. Identify bottleneck

  7. Prioritize improvements based on value and organization goals

Assumptions:

  • Long term averages

  • System operates at steady-state

  • System is full

  • Each process step has a time, rate, and inventory

  • Little’s Law: I=t⋅rI=t\cdot r

Bottleneck: Process activity with longest average time; reduce time at bottleneck to push more units through. Buffers cannot be bottlenecks

Inventory

  • Work in progress: Inventory that has been partially transformed

  • Raw Materials: Inventory that has not been transformed

  • Finished Goods: Inventory that has been fully transformed

  • Inventory Turns: COGS/average inventory

  • Days of Supply: Inverse of inventory turns scaled as days

Increasing throughput

  • Limits need to invest in additional capacity

  • Reduces inventory levels

  • Increases sales volume

  • Improves process flexibility

Competing on Operations: Firm performs better (or at least as well) compared to competition on a specific operational dimension

Process Variability

  • Product variety

  • Structural variation in demand

  • Random variation

  • Assignable (manageable; identify, reduce, and manage)

    • Poor input quality

    • Malfunctions

    • Design flaws

    • Employee training

    • Customers don’t know what to do

    • Delays

Forecasting Principles

  • Methods assume some underlying causal system that existed in the past will persist into the future

  • Forecasts are not perfect

  • Forecast accuracy decreases as the horizon increases

  • Time series data must be first graphed

  • Forecasts for groups of items are more accurate than for individual items

  • Forecasts are continuously updated as new information is available

  • Combination of quantitative techniques and judgement

  • Forecast methods should be reviewed retrospectively for error

  • Steps

    • If you have historical data, graph it

    • Is there a trend?

    • What is the most relevant portion of historical data?

    • Construct a baseline using relevant historical data

    • What factors might influence demand?