ProModel simluation accelerates lean at Generis American Manufacturing Summit
1 like689 views
This document discusses how discrete event simulation can improve and accelerate Lean transformations. It describes how simulation tools have evolved to be more accessible to Lean practitioners. Simulation allows users to test improvements virtually through "Virtual Kaizens" before implementing changes. A case study is presented where simulation identified a 20% increase in throughput and 67% reduction in cycle time for a switch assembly process. The document advocates using simulation to identify high-ROI opportunities and involve operators. It also recommends regularly updating models and linking them to real-time data for more accurate strategic and tactical analysis.
ProModel simluation accelerates lean at Generis American Manufacturing Summit
- 1. 1 Generis – 2015 American Manufacturing Summit Bruce Gladwin, ProModel Corporation bgladwin@promodel.com
- 2. 2 Introductions: Who, What & Why Evolution of (Lean) Process Analytics Simulation-based “Living VSMs” ◦ Understanding the “Physics of the Process” The Virtual Kaizen: your insurance policy Introducing Modeling & Sustaining the Gain
- 3. 3 Bruce Gladwin – fellow pilgrim on the Lean journey 28 years in process analysis and improvement ◦ Hughes Aircraft – Industrial Engineer Role ◦ General Electric R&D – Six Sigma Black Belt ProModel Corporation – Orem, Utah ◦ Simulation Trainer & Consultant ◦ Product Manager, Process Simulation Tool ◦ Currently VP, Modeling Services Team
- 4. 4 What? ◦ How modeling tools, specifically discrete-event simulation, can improve & accelerate your Lean journey Why? ◦ “The times they are a changin” – Bob Dylan ◦ Early modeling tools were not for “the faint of heart” ◦ Advances in simulation technology have put this valuable technology within reach of more and more Lean practitioners Many tools today are “fill in the blank”, no programming required!
- 5. 5
- 6. 6
- 7. 7
- 8. 8
- 9. 9 Learning to See – Lean Enterprise Institute
- 10. 10 Learning to See: © 1998 LEI, Inc., pp. 32-33
- 11. 11 Learning to See: © 1998 LEI, Inc., pp. 78-79
- 12. 12 Value Stream Map ◦ Static model that describes the states of current and/or future processes ◦ Tells what is happening in the process now and what the process could be later ◦ Typically reports on lead time and value added time ◦ Requires general information regarding material arrivals, operation times, labor and equipment availability. e.g. average time = 15 sec Process Simulation ◦ Dynamic model that predicts the behavior of current and/or future processes ◦ Tells why a process behaves the way it does or how it could behave in the future ◦ Reports on throughput, inventory, lead times, VA, NVA and resource utilization ◦ Requires specific information about material arrivals, batch constraints, operation times, labor and equipment uptime & availability, transportation methods and times, and any other capacity constraints. ◦ e.g. Time = T(5,10,25) sec
- 13. 13 Non-functioning Recirculation Loop for empty Tub carriers
- 14. 14 Repairing the Recirculation Loop allowed more WIP which was key to fixing problem
- 15. 15 Throughput Up 37% !! WIP also Up 37% !!
- 16. 16
- 17. 17 Interdependencies Variation SIMULATE Use Simulation When You Cannot Ignore Variation and Interdependencies Static Tools SPREADSHEET ANALYSIS MONTE-CARLO SIMULATION (Crystal Ball) OPTIMIZATION ALGORITHMS (LP, MIP)
- 18. 18 Using “Averages” can be dangerous. How much Risk can you afford to take?
- 19. 19 Trying to optimize a complex mfg system is like trying to solve the Rubic’s cube. Simulation can help understand the trade-offs!! Maximize Throughput Reduce Cycle Time Reduce Inventory Minimize P&E Interdependencies
- 20. 20 Exercise 1 … from Process Simulator Basics Training Course Activity Worker Work Unit (Exit) 5 minute Process Time 100% Resource Availability Deterministic Example Activity 2 Worker 2 Work Unit 2 (Exit) U(5,1) minute Process Time 100% Resource Availability Stochastic Example Activity 3 Worker 3 Work Unit 3 (Exit) U(4,1) minute Process Time 80% Resource Availability Stochastic Example with Uncertainty Activity 4 Worker 4 Work Unit 4 (Exit) U(4,1) minute Process Time 80% Resource Availability Assembly Opn Stochastic Example with Uncertainty and Interdependency Activity 4A Worker 4A Work Unit 4A BA C D Arrival Rate = 1 unit every 5 minutes Identical for every arrival stream No variation
- 21. 21 Results Are Striking! 30 Replications: 40 hr Comparison of System A to System D Throughput: - 4.6% Avg Queue: + 14x Avg Cycle: + 15x
- 22. 22
- 23. 23 A Virtual Kaizen is simply using simulation first to better understand the “Physics of the Process”… Then design the improved process, to produce the desired outputs, within expect levels of variation… All in a risk free environment ◦ No interruptions to the existing line ◦ No need to make physical rearrangements (yet) ◦ Ability to ask “What If”
- 24. 24 Founded in 1975 by former Westinghouse engineers Ownership: privately held by founders Location: Pittsburgh area Employees: 150 total (50 salary, 100 hourly) Products: ◦ Mechanical Switches for Electric Utility Grids ◦ Automatic Distribution Motor Operators (ADMO) Allows remote control of switches by utility personnel
- 25. 25 Vertical Switch V2-C (copper) Center Break Switch CB-A (aluminum) 6’ 6’ Switches = 50% Revenues
- 26. 26 Privately held, debt-free, profitable business with little competition from low cost countries (for now). Capacity constrained. The market would buy every additional unit that they could produce. Each additional unit sold will contribute about $5k- $10k to their bottom line. Want to maximize 1st shift capacity before adding additional 2nd shift operations. Considering 30,000 sq ft expansion to their existing plant, at a cost of $1M.
- 27. 27 Eliminate or Reduce the Risk through a “Virtual Kaizen” using Simulation Technology.
- 28. 28
- 29. 29 Value stream mapping Work standardization Point of use storage Single Piece Flow Visual management (Subassembly Feeder Lines) Kanban (pull) production (Subassembly Area) Setup reduction (Fabrication Area) Total productive maintenance (Fabrication Area)
- 30. 30
- 31. 31 Simulation proved to be an accurate tool for modeling the current state Vertical Switch Assembly process A Future-State model using Lean Production concepts proved to be far more efficient ◦ 20% improvement in product throughput! ◦ 67% reduction in average unit cycle time! ◦ 50% reduction in floor space requirements! ◦ Elimination of need for 2nd shift!
- 32. 32 Identify the best opportunities ◦ Good ROI potential Every additional unit produced can be sold ◦ (Fairly) Low Risk Use a Sensei with years of modeling experience Involve all stakeholders, especially line operators whenever possible Be patient and allow your team time to develop modeling skillset. Make modeling a part of every major process change.
- 33. 33 Update the models regularly so they are always available to predict the future ◦ More & better data results in better decisions Link to real-time data sources ◦ More & better data results in more accurate results ◦ Strategic Analysis becomes Tactical Analysis
- 34. 34 Generis – 2015 American Manufacturing Summit Bruce Gladwin, ProModel Corporation bgladwin@promodel.com