Impacting the Bottom Line. Cost Reduction: Planning & Control. Bottom Line (costs) A = Costs (what we expend) B = Revenue, sales (what we generate)

Transcription

1 Cost Reduction: Planning & Control Break Even Point & Decision Tree Analysis (l.u. 4/2/13) Impacting the Bottom Line Calls for an understanding of: Variation Waste & Value Investments (time, capital, resources, ) - ROI Design Relationship of activities (sequence, connection) - TIME Bottom Line (costs) A = Costs (what we expend) B = Revenue, sales (what we generate) Below intersection = LOSS LOSS Above intersection = PROFIT Bottom Line (costs) 1. Fixed (rent, property taxes, loan payments) can be estimated in advance, year or more 2. Variable (labor, contracts, utilities, travel) difficult to estimate, apply ceiling/limits for further expenditures (what we target in cost reduction) Direct (labor, materials) Overhead Indirect (benefits, purchasing, marketing, management, travel) Facilities (rent, maintenance) Taxes/insurance) (Angus, Gundersen & Cullinane, 2000) (Minty, 1998, p. 98) Bottom Line The comparison of project income versus project expenditures is referred to by business and industry executives as the bottom line (Angus, Gundersen & Cullinane, 2000, p. 186) Bottom Line: Investments Break Even point? Capital cost of Machine + Operational Costs Face value profit if the machine lasts 10 years Want Decrease Slope Want Increase Slope (Angus, Gundersen & Cullinane, 2000, p. 186) (Minty, 1998, p. 104) 1

2 Investments: Break Even Point Capital cost of Machine + Operational Costs do not cover everything (DOUBT, maintenance, time value of money, scrap value) Maintenance Overhaul Uncertainty Investments: 10 Year Projection However, data conflicts (so we use decision tree) with probability Senario1: $80k startup, t = 10 years (10 x 2000 hrs), $5/hr operation/maintenance, $0 scrap value. Cost = $180,000 Senario2: $80k startup, t = 10 years (10 x 2000 hrs), $5/hr operation/maintenance, $10k scrap value. Cost = $170,000 Senario3: $80k startup, t = 8 years (8 x 2000 hrs), $5/hr operation/maintenance, $40k 2 years labor. Cost = $200,000 Steeper Increase Wages Senario4: $80k startup, t = 5 years (5 x 2000 hrs), $5/hr operation/maintenance, $80k new robot. Cost = $260,000 (Minty, 1998, p. 104) Investments Decision Trees From Data/Research Startup, Maintenance $5/hr, & Scrap Value Probability x Cost Questions? If you are a manager (project, site, operations, engineering, supply chain): What common variable significantly impacts bottom line? What are decisions based on for any project? What activities do we improve/eliminate? How do we determine where to deploy resources? Break Even Point weighing risks (Minty, 1998, p. 107) Cost Reduction: Planning & Control Project Flow (Theory of Constraints) (l.u. 4/2/13) Fallacy = Fix Everything that is Waste! 1. PROBLEM! Diminishing returns: Traditional cost accounting attempts to maximize the utilization of resources and work centers even if they build inventory that is not needed and even if the work centers are not bottlenecks (Schroeder, 2008, p. 301). When companies spend $ if this does not positively impact the bottom line, then it is waste 2. Most companies use lean tools to make point based improvements that do not impact the bottom line this is not lean. (Kevin J. Duggan, personal communication) PEOPLE MUST UNDERSTAND THE FLOW OF OPERATIONS AND HOW PROCESSES INTERRELATE WITH ONE ANOTHER 2

3 Theory of Constraints (TOC) Decisions are based on money (hopefully long term)! Businesses exist to make $ if they don t make $, they cease to exist (unless the government, AKA you, bail them out!) Eliyahu Goldratt, author of The Goal Where is the Bottom Line? If you are a manager under time/resource constraints, how do you determine WHERE to make improvements? HOW do you identify the time that impacts the BOTTOM LINE? 1. Identify bottleneck in system 2. Seek to improve capacity of bottleneck to improve throughput (reduce setup time/changeover time, better scheduling, 24 hour operation, better workforce policies, reducing inventory, etc.) (Schroeder, 2008) TOTAL SYSTEM CAPACITY CAPABILITY ($) TIME Gantt Charts Bar Chart oldest and most frequently used chart for plotting work activities against time (Angus, Gundersen & Cullinane, 2000) Quickly interpreted Gantt Charts Inevitably, schedules become off track and adjustments must be made (with perhaps little understanding of consequences). (Minty, 1998, p. 67) (Minty, 1998, p. 69) Gantt Charts Used to allocate resources (e.g. Scheduling people) projecting days/weeks in advance Best used for simple/few operations Gantt Charts Disadvantages: they do not display the relationships between activities. If one of the activities is delayed, the chart does not convey whether this will affect the beginning of some other activity. For example, in the building of a house, if the plumbing is delayed, can the exterior painting and siding continue as planned? You cannot tell from a Gantt chart. (Minty, 1998, p. 68) So, Gantt charts do not illustrate how disruptions in a schedule result in disruptions in FLOW OF OPERATIONS (delays)! Gantt charts are justified for projects where the activities are not highly interconnected or for small projects (Schroeder, 2008) (Minty, 1998, p. 68) 3

4 Simple Question You have 3 suppliers providing 3 parts for an assembly project (all 3 parts must be received before assembly begins) Delivery: Two suppliers = 1 periods, One supplier = 6 periods Where do you invest your time/money? What company are you concerned with most? Network Charts PERT Chart Program Evaluation and Review Technique (PERT) 1950 s (Minty, 1998; Schroeder, 2008) Polaris nuclear submarine project (1 st one ever), 3000 contractors, 2 years ahead of schedule (Schroeder, 2008) Common Mapping Approaches: PERT VSM (Minty, 1998, p. 71) Network Charts PERT Chart Reveal sequence of events in path Reveal simultaneous paths precedence relationships are explicitly shown (Schroeder, 2008) Longest path is the Critical Path; delays are detrimental along this path (Minty, 1998). Elsayed and Boucher (1985) cite a study that found as many as 80% of 400 construction firms were using the critical path method (Angus, Gundersen & Cullinane, 2000, p. 179). (Critical Path Method) estimates a single time value estimate for completion, while PERT Methods calculate uncertainties and probabilities (Angus, Gundersen & Cullinane, 2000). PERT/ Supporting Software To name a few: Omnilab Microsoft Visio Microsoft Project Pert Chart Expert PlanBee Primavera SuperProject Scitor PS Suite ProChain Solutions RFFlow PERT Chart Paths CP Critical Path = Greatest elapsed time = ACFHJLP = 20 days Example Problem A 1. Determine the shortest time (days) to complete the project. 2. How many person/days are needed to complete the project? 3. What day will activity H-M require resources? (Minty, 1998, p. 75) (Minty, 1998, p. 71) 4

5 Example Problem A Problem 1: Network Chart Calcs 1. What are the events along the Critical Path? 2. How long to complete this project? 3. How long to complete the project if Activity E I takes 3 days longer? 4. How long to complete the project if Activity K N takes 1 day longer? 5. How long to complete the project if Activity C G takes 5 days longer? 4. What day will activity L-N require resources? (Minty, 1998, p. 75) 5. If K-L requires 2 days more than projected, how long will it take to complete the entire project? 6. Construct a Gantt chart (schedule) for the project. (Minty, 1998, p. 87) Project Delays: Management Options Impact Critical Path NO 1. Note slack taken. 2. Verify impact on other paths. 3. Root Cause Analysis: verify what went wrong & why. Then Design out problems for future projects. Impact Critical Path YES 1. Contact the customer verify their impact 2. Reallocate resources to other phases of project (overtime, extra shifts, temporary workers, etc.) 3. Root Cause Analysis: verify what went wrong & why. Then Design out problems for future projects. Problem 2: Network Chart Calcs 1. Draw a Network Chart (ACTIVITY ON NODE) with the following Parameters: Z can not be completed until A and B occur A cannot occur until C and D occur B cannot occur until E, F, G occur E, F, and J cannot occur until H occurs G cannot occur until I occurs I cannot occur until J occurs C, D and H cannot occur until K occurs 2. Each activity takes 3 periods What is the completion time? 3. What path do you target for making improvements? 4. If Activity C is delayed by 4 periods, what is the earliest completion time? Network Chart Calculations PERT: Estimated Time Some activities in projects have uncertainty due to variables, estimates made from statistical data Times for activities: OPTIMISTIC (we hope), PESSIMISTIC (1/100, with delays), and PROBABLE (highly likely) Time Estimates t = single time estimate o = optimistic time n = probable time p = pessimistic time t = (o + 4n + p)/6 Estimated completion time A-B: t = (5 + 4(6) +10)/6 = 6.5 periods Optimistic Probable Pessimistic t (Minty, 1998, p. 78) 5

6 PERT: Estimated Time First Calculate all probable times between each event Second Determine the CP Third Add probable times OR PERT: Estimated Time t e = (o + 4n + p)/6 TIME ESTIMATE Earliest A G = 13 periods, Latest A G = 24 periods t e = (13 + 4(17) + 24)/6 = 17.5 periods (total time estimate) OR t e = = 17.5 periods (total time estimate) First Add all o, n, p about the CP Second Apply the standard PERT Single Time Estimate Formula (easier) (Minty, 1998, p. 78) PERT: Estimated Time Problem 3: Estimated Time Given the data, calculate the probable time for completion (Angus, Gundersen & Cullinane, 2000, p. 183) (Minty, 1998, p. 89) Problem 4: Calculating Slack with Problem 4: Calculating Slack with Earliest Start ES(X) = 0 [starting activities] Activity time t(x) Earliest Finish (EOT) EF(X) = ES(X) + t(x) Latest Start LS(X) Latest Finish LOT LF(X) = LS(X) + t(x) Earliest Occurrence Time EOT Latest Occurrence Time LOT Slack = Amount of time an event can slip without affecting project completion date (Angus, Gundersen & Cullinane, 2000, p. 178) 6

7 Problem 4: Calculating Slack with Problem 4b: Calculating Slack with (Schroeder, 2008, p. 323) (Angus, Gundersen & Cullinane, 2000, p ) Problem 4b: Calculating Slack with Calculating Variability & Probability NOTE: Experience shows that time estimates often exceed the most likely time or best estimate in project activities because people tend to be overly optimistic in their time estimating (Schroeder, 2008, p. 325). Variance of the complete project can be computed along the critical path using the time estimate and ACTIVITY VARIABILITY. It is likely that each activity will experience some variability from the time estimate. ACTIVITY VARIANCE (Schroeder, 2008, p. 324) Problem 5: Calculating Variability/Probability Q: Given the history of the following process, what is the probability that the project will be complete within 12 time periods? Problem 5: Calculating Variability/Probability Construct chart Compute estimated times Compute time variances Determine CP Time 1. Construct chart 2. Compute estimated times 3. Compute time variances Then 4. CP Activity On Node (AON) Technique (Schroeder, 2008, p. 326) (Schroeder, 2008, p. 326) 7

8 Advantages/Disadvantages Network Diagrams: Illustrate disruptions in flow of activities Show interrelationships between/among activities Reveal the bottom line [CP](where improvements are significant) helps prioritize expenses Do not differentiate time (value vs waste) as in VSM May be complicated by statistics if management is not trained or unwilling Scale of some large projects may be difficult to display Resources expended may not justify benefits due to short project duration References Angus, R. B., Gundersen, N. R., & Cullinane, T. P. (2000). Planning, performing, and controlling projects: Principles and applications. Upper Saddle River, New Jersey: Prentice Hall. Minty, G. (1998). Production planning and controlling: A problem based approach. Tinley Park, Illinois: Goodheart Willcox. Schroeder, R. G. (2008). Operations management: Contemporary concepts and cases. New York: McGraw Hill Irwin. 8