MGSC 395 Exam 2 Study Guide
MGSC 395 Exam 2 Study Guide MGSC 395
Popular in Operations Management
verified elite notetaker
Popular in Business, management
This 11 page Study Guide was uploaded by Rachel Whitbeck on Sunday March 27, 2016. The Study Guide belongs to MGSC 395 at University of South Carolina taught by Pearse Gaffney in Spring 2016. Since its upload, it has received 133 views. For similar materials see Operations Management in Business, management at University of South Carolina.
Reviews for MGSC 395 Exam 2 Study Guide
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 03/27/16
MGSC 395 MW 5:30 Wednesday, February 17, 2016 THE GOAL DISCUSSION - Alex Rogo: main character, newly appointed plant manager - Jonah: Alex’s old college physics professor, mentors Alex through turning the plant around Throughput: the rate at which the system generates money through “sales” Inventory: all the money the system has invested in purchasing things to “sell” o Contrary to what we teach even today, inventory is money that’s tied up, it’s more of a liability Operational expense: all the money the system spends in order to turn inventory into throughput - The goal is to make money - “Herbie” or the bottleneck sets the pace. o Dependent event- everyone is dependent on Herbie to get to the campsite o Statistical fluctuations- happen when the boy scouts spread out due to all their different paces Theory of constraints o Identify the system’s constrains (--> NCX-10 machine in The Goal) o Decide how to exploit the constraint (--> stagger lunch breaks so the constraining machine is never idle) o Subordinate everything else to that decision (--> Red tags are top priority, and green tags come second) o Elevate the system’s constraints (--> bring back the old Zmegma machine to help the NCX-10 and increase capacity// or outsource o If you break a bottleneck, do not let inertia set in; go back to step 1!! Monday, February 22, 2016 Bottleneck identification o It has the highest total process time per unit o It has the highest average utilization and total workload o Drum – Buffer – Rope Drum: sets the pace; pacemaker Buffer: inventory put in front of the bottleneck Product Mix (example 5.2 and 5.3) (Use a spreadsheet for the homework) Traditional Method o 5 workstations: V, W, X, Y, Z o 40 hour work week o Product Demand Price R.M. Contribution Margin A 60/week $75/unit $10/unit $65/unit B 80 72 5 67 C 80 45 5 40 D 100 38 10 28 o Product Flow and process time A V(30min) --> x(10min) --> Y(10min) B X(20) --> Y(10) C W(5) --> X(5) --> Y(5) --> Z(5) D W(15) --> Y(5) --> Z(10) o Priority based on contribution margin B A C D o Workstation Mins available Mins after 80 B*Mins after 60 A at start V 2400 2400 600 W 2400 2400 2400 X 2400 800 200 Y 2400 1600 1000 Z 2400 2400 2400 *B is top priority, and 80 is the demand for B. Multiply the number demanded by the number of minutes required at the workstation in the flow and process time chart According to theory of constraints, manage the bottleneck o In this problem, the bottleneck managed us instead, so there’s another method: the bottleneck method The Bottleneck Method o Calculate the CM per minute at the bottleneck o Product Demand Price Resource CM CM/unit Materials @ BN A 60/week $75/unit $10/unit $65/unit 6.50 B 80 72 5 67 3.35 C 80 45 5 40 8.00 D 100 38 10 28 N/A* *Doesn’t go through bottleneck o New priority based on contribution margin per minute at bottleneck D C A B o Workstation Mins available Mins after 80 D Mins after 60 C Etc… at start V 2400 2400 2400 W 2400 900 500 X 2400 2400 2000 Y 2400 1900 1500 Z 2400 1400 1000 Wednesday, February 24, 2016 Line balancing: The assignment of work to stations in a line process to achieve the desired output with the fewest workstations possible Steps o Step 1: Separate all the work into elements o Step 2: Get time estimates for each element o Step 3: Identify the immediate predecessor for each element o Step 4: Calculate the line’s cycle time C = 1/r , where r is the demand rate o Step 5: a) Calculate the theoretical minimum number of workstations. o TM = ∑t/c , which is the sum of the times for each machine over the cycle time you got in step 4 o If TM is 4.61, you round up and that new number is called “n” o This means that 90 seconds (the cycle time in the example in class) are available at each of the 5 workstations. b) Balance the line. o Bar graph like in class o Step 6: Calculate idle time, efficiency, balance delay. Idle time- just add together the time not used at each workstation % Efficiency = (∑t/nc) x 100 o See notebook paper for full example In this class, we can only do steps 4 – 6. The term balance delay isn’t actually used in the real world. It’s just the percentage of inefficiency Monday, February 29, 2016 Linear programming L-P o A method to allocate scarce resources among competing demands Is an optimization methodology For us, this means we will try to either maximize or minimize something o The objective function- a math expression of the thing we will maximize or minimize o Decision variable- the things we can control for which we are trying to find the optimal value o Constraints Hard limits that impact our ability to choose something about our decision variables Have to be <, =, or > o Steps to set up an L-P 1. Define the decision variables 2. Write out the objective function 3. Wrote out the constraints o RHS- right hand side o POM- defaults to 0, don’t forget the 1 Wednesday, March 2, 2016 Lean production systems Notes are mostly on powerpoint o 7 (or 8) wastes o Remember TIM WOOD o Lean house o Lean production tools o Takt time- run your machine 24/7 to get your money’s worth, but don’t run it at all of you don’t have enough demand Available time÷ demand o Kanban- control production flow Limits production Means card or signal Supermarket- backwards E looking thing When someone buys a thing, that’s the signal to make another one of those things Kaizan= continuous improvements o Poka yoke- error proof o 5S- sort, set in order, shine, standardize, sustain o Six sigma o 99% good is not good enough o 99.99966% is o Bell curve will move- shift happens Monday, March 14, 2016 Project management Planning projects o 6 p’s Work breakdown structure (WBS) o Tree diagram- does not have to be symmetric or sequential o Two primary scheduling methodologies Activity on node diagram Once you have your chart of activities, predecessors, and time estimates, you can make this chart, which is literally the name of the activity on a circle, or node. You can calculate all the possible paths and their durations o We are always interested in the one that takes the longest amount of time (called the critical path) o This is called the critical path method (CPM) PERT- Program Evaluation and Review Technique Only difference between this and CPM is that you estimate the longest and shortest times for each activity and use statistical methods to know the possible variations o Gantt chart Popular in Microsoft projects Bar chart Wednesday, March 16, 2016 Hospital project o ES= early start, EF= early finish Forward pass (see photos) Begin at time=0 (or immediately) Move forward “Bigger number wins” Only looks at “early,” because it shows the earliest time that each event can start and finish o LS= late start, LF= late finish Backward pass Begin at critical path duration at the end Move backwards through the network diagram “Smaller number wins” Only looks at “late” o Finding slack Once you’ve done forward and backward passes, go through either subtract early finish time from late finish time, or subtract early start from late start. S= LF – EF, S= LS – ES You need to find slack to know how much wiggle room you have If something on the critical path looks like it’s slipping, you can move people from a non-critical activity with slack over to the critical path activity so that you don’t actually lose any time o Cost time tradeoff (minimum cost schedule, or crashing the project) Go to each manager of each project and tell them to do better than the Normal Time and Normal Cost They will come up with shorter times, but higher costs (Crash Time and Crash Cost) Max time reduction: NT – CT Cost to crash per period: (CC – NC)/(NT – CT) Process Identify the critical path Find the cheapest activity from the critical path (cost to crash per period) Crash for as much time as you can until: o Another path becomes more critical o No more time can be taken o Savings not there Go back to first step and repeat First pass Critical path at 69 weeks (hospital example) Activity J has the cheapest cost to crash per period at $1,000 per week o Saves 3 weeks (put this into the other possible paths to make sure none are becoming more critical than the original critical path of BDHJK) Cost: 3 weeks at $1,000 per week = $3,000 o Gross savings: indirect --> 3 x $8,000/wk = $24,000 Penalty--> 3 x 20,000/wk= $60,000 Total of $84,000 Second pass Repeat what you did in the first pass with the next cheapest activity Third pass If you have two critical paths now because of the time you took out, like now we have BDHJK as well as ACGJK, you have to take time out that would influence both paths, not just the new one Fourth pass Do B and C simultaneously (two weeks out of an activity in each of the two critical paths POM o Module: o Method: Crashing o Watch out for the word AFTER at the top (for hospital problem, you’d put after week 65 Monday, March 21, 2016 Time series models (all models are wrong. Accept it. We want to figure out how wrong are they) o Graph with time as x and demand quantity as y o Your data can show as horizontal, a trend (diagonal up or down), cyclical (wave shape), or seasonality (horizontal then spike) o E t D --tF (trror at time t = actual demand minus forecasted demand) o Get the sum (∑ CFE= cumulative forecast error) and the average (x bar = average bias) To get rid of the negative signs in the errors, take the absolute value. Then x bar is MAD (mean absolute value) Or you can square the error. Then x bar is MSE (mean squared error) Or you can take the percent error, which is the percent of the absolute error. Then x bar is MAPE (mean absolute percent error) The percent error is the error divided by the actual demand Time series methods o Option 1- Naïve forecast F =4 3 Forecast for month 4 is the demand from month 3 This doesn’t take anything into account, like seasonality o Option 2- Simple moving average Have to know length of the moving average Example: you’re given the actual demands for months 1 through 3. Forecast for the 4 month. o F =4(D +1D + D2)/n 3 F5= (D +2D + 3 )/n 4 o You don’t use D because you only use the 1 last three months in your average o Option 3- weighted moving average Assign a weighted percentage to each month Just multiple the weight by the demand for each month when you’re adding them in the formula F4= %D + 3D + %D 2 1 o Option 4- exponential smoothing Requires this period’s forecast, this period’s actual demand, and a smoothing parameter, or ∂ (alpha) (a number between 0 and 1) F = ∂D + (1--∂)F t+1 t t Wednesday, March 23, 2016 Forecasting o Casual relationship on a scatterplot comparing sales quantity and advertising $ o The fitted line on the scattered plot is the regression line. Y=mx+b This just means it’s the best line to fit the data with the least squares error (square to eliminate negative numbers, like last class) o Three measures (statistics) to assess “goodness of fit” R--> the sample correlation coefficient --> measures the strength and direction of the relationship You can tell if it’s downward/negative (r= -1) or upward/positive (r=1) or if there’s no relationship (r=0)j R can be from -1 to +1 R --> the sample coefficient of determination --> measures the amount of variation explained by the regression (between 0 and 1) Closer to 1 implies strong relationship with most variation explained Sxy--> the standard error of the estimate Like the fat pencil test (the pencil covers up most of the dots. You want them to be tightly clustered around the line) o PGA- best advice ever. Practical, graphical, analytical o Fitting the line- linear regression Start with data Access the casual relationship Use software to “fit” the line? Identify the regression equation Check if your line is a good fit o Basically just look at the answer the computer gives you and see if it’s even logical If you’re given data and it says 2.5 was spent of advertising, it’s probably not $2.50. It’s probably $250,000 Correlation does NOT imply causation Monday, March 28, 2016- test review Set up a linear program with math o Find decision variables o Write objective function o Write out constraints o (All in a form that uses the decision variables) HW problem 3 on project management o Crashing, aka cost time tradeoff, aka minimum cost (and) schedule Goal is to reduce time while also reducing cost For your chart, use the labels below for each column Activity| Normal time| Immediate predecessor| Normal cost| Crash time| Crash cost| Max time reduction (normal time minus crash time)| Cost to crash/period (cost to crash minus normal cost, divided by the normal time minus crash time) o Remember the process: Identify the critical path Identify the activity on the critical path with the lowest cost to crash per period (CC-NC)/(NT-CT) Crash that activity until o You can’t take any more time o Another path becomes more critical o Or you aren’t getting any more savings out of it Go back to the first step and do it again with the next cheapest activity Slack o Use a forward pass to find the “earlys”- start and finish Start at time = 0 For example, if activity A is first and takes 4 weeks, the earliest it can start is 0 and the earliest it can finish is 4. Bigger number wins You’ll end at the critical path’s duration o Use a backward pass to find the “lates”- start and finish Start at the end, which will be at the critical path’s duration For example, if the CP duration is 18 weeks, and the last activity (F) takes 7 weeks, then the latest F could finish is 18 weeks and the latest it could start is 12 weeks. Smaller number wins this time You should end at the beginning with 0 as the late start time o To calculate slack, do either LF minus EF, or LS minus ES. You really only need to calculate either the finishes or the starts For the goal: o Know the theory of constrains (look in the textbook) o Also would be beneficial to look at online notes about the book Line balancing (geared specifically toward assembly lines) o 6 step process (but steps 1-3 are given to you) Steps 1-3 are to identify all the work, break it up into elements, and get time and precedence relationships between activities Step 4: calculate the line’s cycle time (how often should a car come off the assembly line). C=1/r where C= cycle time and r= demand rate Step 5: Calculate the theoretical minimum number of workstations T.M. = (∑t)/C where TM is theoretical minimum, ∑t is the sum of all the times, and C is the cycle time Always round up for this number (you can’t have a fraction of a machine or workstation) If you have one sucker who doesn’t fit into the theoretical number of workstations, then you have no choice but to add another workstation.
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'