Let’s Begin …

IEM 5303Advanced

Manufacturing Systems Design

2000 John W. Nazemetz

Slide 1

Welcome to Week 10 Discussion

Agenda for Tonight:

Review of Exam

Discussion Questions from Discussion Leaders

Other Questions from Students

Let’s Begin …

IEM 5303Advanced



Slide 2

Question 1 --

1. (25 pts.) Do CIM systems that employ Generative Computer Aided Process Planning benefit from the simultaneous use of Group Technology concepts?

Yes/No

Exam Review (1a)

IEM 5303Advanced



Slide 3

Question 1 -- cont. If so, how and/or under what conditions? (25 of 25 possible)Similar shapes/designs -> similar routingsOperational Efficiencies (families, batching in

families – recalculate families each period)Reduction in part variation, increase volume

of common partsReduction in tooling with reduction in parts

variation, tooling cost with volume increase

Exam Review (1b)

IEM 5303Advanced



Slide 4

Question 1 -- cont. If not, why not (20/25 max available).

Generative CAPP, re-determines routing each time based on technical and economic feasibility so no permanent families established/establishable

BUT -- GT makes designs similar, so algorithmically derived routings are similar and other benefits are derivable.

Exam Review (1c)

IEM 5303Advanced



Slide 5

Exam Review (2)

Question 2 --(25 pts.) Define and explain Quality Function Deployment. List

and discuss its weak points. Define and Explain (15 pts.)

Process to document and assess products with respect to an ordered and weighted set of customer requirements, technical specifications, the correlation of customer requirements and technical specifications, comparison to competitors, required quality (tolerances) of part, production methods.

Weaknesses (10 pts.)Ability to gather customer

requirements/competitor dataAbility to translate requirements to specificationsAbility to define correct weights/economicsCan make week analysis “look” stronger than it is

IEM 5303Advanced



Slide 6

Exam Review (3)

Question 3 --Consider the various models proposed in the text and in

lecture for developing part families and manufacturing cell groupings. List, briefly define, and evaluate three (3) models/methods for cell and/or family formation (i.e., explain the philosophy/objective of each algorithm/procedure, the steps of each algorithm/procedure, and discuss its strengths and shortcomings). (8 pts each plus 1)Machine Component Group AnalysisProduction Flow AnalysisRank Order Clustering (King)GT Coding and Pattern RecognitionSingle-Linkage Cluster Analysis - Similarity Coefficients

IEM 5303Advanced



Slide 7

Exam Review (4a)

Question 4a --(17 pts.) Consider Robots and Numerical Control Machines.

Explain the similarities and differences between the two. Similarities

Same basic technology, sensorsSimilar programmingSimilar function (move parts or tools in space)Similar concepts – open/closed loop, pt to pt,

continuous path, multi-axis, multiple tools/parts

DifferencesRobots less precise, accumulate error while NC

more precise, axis error is independentRobot generally used for ergonomic and economic

reasons, NC generally used for technological reasons

IEM 5303Advanced



Slide 8

Exam Review (4b)

Question 4 --(8 pts.) Consider Robots and Numerical Control Machines.

Briefly explain the role/use of each in FMS/Cellular Manufacturing under conditions of a) high product (design/shape) variability and low part volume (sales) as well as b) high part volume (sales) and low product (design/shape) variability.

Role – high product variability/low volumeRobot – allows flexibility via general purpose hand,

teach or low-cost/time programmingNC – allows precision manufacturing via

reprogrammingRole – low product variability/high volume

Robot – allows labor replacement, repetitive motionsNC – provides an alternative to custom, single

purpose equip.

IEM 5303Advanced



Slide 9

Exam Review (5a)

Question 5 --(25 pts.) Briefly define the life cycle and product concepts

of Chris Vaughan. (15 pts.)Life Cycle Concepts:

Design, Produce, Operate, Service, DisposeAll products have this life cycleIterative process (Less -> More detailed)Applies across Supply Chain (all Levels)

Product Concepts:Almost everything is a productProducts interact (Operate facility (product of “design and produce facility” is produces product

IEM 5303Advanced



Slide 10

Exam Review (5b)

Question 5 --(25 pts.) Briefly define the life cycle and product concepts

of Chris Vaughan. Comprehensively discuss how his definition/concepts apply/do not apply in the practice of Concurrent Engineering. (10 pts.)Phases of life cycle define major inputs to concurrent

engineering (design must consider production, operation, service and disposal of product)

Iteration of design Design document(s) (product) –

output/input of design facility, iterates from conceptual design to “final” design to redesign to re-redesign, …

Stages of design correspond to (design) supply chain

IEM 5303Advanced



Slide 11

Discussion Questions (Jeff Short)

IEM 5303Advanced



Slide 12


What is the difference between JIT, MRP, and Lean Manufacturing?JIT – minimum (zero) inventory goal

Absorbs fluctuations/disruptions via excess capacity (labor and machines)

MRP – maximize utilizationAbsorbs fluctuation by planning/scheduling

Lean Manufacturing – a manufacturing philosophy of efficiency which, when implemented, shortens the time between customer order and factory shipment by eliminating waste. "Uses less of everything compared with mass production”

Focus on efficiency/productivity of process

IEM 5303Advanced



Slide 13

Discussion Questions (Jeff Short)Chapter 10 is about MRP and Chapter 11 is concerned with JIT. These

seem to be diametrically opposed views of production planning and control. What is the current state of research to resolve the philosophical disagreements and reach a more practical operating philosophy?

Diametrically opposed in theoretically pure form –In practice, they blend –

Balance ability to predict and preplan fluctuating vs. constant stable system (production, customer, …)

No smooth/known demands (Seasonality, competitors)

No smooth production (machine/managerial breakdown)

No 100% availability (temporary demand exceeding capacity, rush orders)

Inventory to deal with unexpectedResearch using each method criteria (utilization/inventory)

IEM 5303Advanced



Slide 14


Can JIT exist in practice for a manufacturer with high variance in demand?

Yes, but if variance in demand is high, a large amount of excess production capability must be available to meet demand (on-time). As this cascades through a system, more interchangeable, cross trained workers needed with resultant higher labor costs.

IEM 5303Advanced



Slide 15


I understood the Wagner and Whitin Algorithm produced this property: either the inventory carried to period t+1 from period t must be zero or the production quantity in period t+1 is zero.

Question 6 asks the student to begin with an inventory, obviously a violation of the W-W algorithm. I felt there were several ways to deal with this problem, all of which modified the algorithm. Which way would you suggest?Reduce demand in first period(s) to reduce inventory to

zero.Is there a stochastic or fuzzy representation of the W-W

algorithm??Yes, simulation/modified Newsboy problem

“goodness of solution” is determined by ability to predict period demands/distribution of demand.

IEM 5303Advanced



Slide 16


Both W-W and Naidu & Singh treat the variable of h, holding costs, as a well-know entity. I have never worked for/with a company that could fix the value of h for any product under all storage conditions. How is h handled in your experience?Via “Best Estimate” Usage

V = value of part/product ($)h = V x (interest rate +inflation rate +

insurance rate + spoilage rate + pilferage rate + management rate)+ Size x (cost of space)

All rates in ($/$inv), Cost of Space in $/unit volume

Can stratify inventory (Pareto-type Analysis)

IEM 5303Advanced



Slide 17

Discussion Questions (Mike Blanton)

IEM 5303Advanced



Slide 18


Problem 10.7 was difficult to do. It followed example 10.4 which was continued from example 10.3. The examples seemed to leave information out, under the assumption that they were easy to follow. For me and I assume other students this was not the case. Could you please fill in the blanks.I’ll try on the next slides.

IEM 5303Advanced



Slide 19


In example/problem, an LP model is solved.For 10.7 –

One must first calculate the demand (dt)In period one dt = 50 + 60 + 70 + 80 = 260

unitsIn common units Xt = 2(50+70) + (60+80) =

380 hoursThis will mean k = 1 hour/unit for 380 unitsPerform similar calculations for other

periods.

Then set up LP problem:

tfthtstitutotwtx FcHcIcIcUcOcWcXcMin

IEM 5303Advanced



Slide 20


Subject to

With Wo=240, k=1d1=380All 0 subscripts droppedNon-negativity

tttttttt FHIIUOWXMin 40245005604024120

tttt

tttt

tttttt

WkXUOFHWW

IIIIXd

1

11

IEM 5303Advanced



Slide 21

Discussion Questions (Class)

IEM 5303Advanced



Slide 22

Discussion Questions (Class)

Other Questions as raised

IEM 5303Advanced



Slide 23

Discussion Session 10

End of Class

Have a Good Week and I’ll see you next time!

Let’s Begin …

Documents