Chapter 6: Value Engineering

1
Ch-29 PROGRESS CURVES or
PRACTICE MAKES PERFECT
2
Progress (Learning) Curves
Individual - Location for
Learning Organization
Log-Log
Concept Det. A Curve - Quantifying
Improvement Typical Values
Progress Typical Learning
Values Cost Allocation Scheduling -
Applications Alternatives Evaluation Acceptab
le Days Work Incentives
3
Progress Curves (cont)
General Assumption in Industry Time/unit
cte vs cummulative units (warehouse, storing)
produced. It is a source of error in
Scheduling, Evaluation of Alternatives, fair pay
and incentive pay Individual Learning
(performance improvement) John is practicing his
service in tennis. He has a hammer racket, and
Wilson 7 balls. From sets of 10 balls, in the
first round 5 make it to the other side. Then 6,
7, ... Eventually John got them all OK. He
further improve by putting the balls where it
hurts. All other variables were kept the same.
The court, the high of the net, the racket, the
balls, etc. His performance improved because he
got a better grip, put his fit together, through
the ball higher, get a better arm swing
(extension), warm up, etc.
4
Progress (Learning) Curves
In his job John does the same. Same tools, same
product, he turns out more widgets/hr, he gains
more experience. Reasons for Improvement
Better eye-hand coordination, fewer mistakes,
reduced decision-time, Organization Learning
or Manufacturing Progress Changing Product
design, tools and equipment, work
methods Individual Organization
learning John might change raquet, balls, tennis
shoes, etc. Other Cases Coffee serving,
Airplane Fab. parts, Subway
5
Progress (Learning) Curves

• Organization Learning
• Operator learning with existing technology
• Influence of new Technology Solid state
  electronics better TV sets, Laptops with higher
  resolution speed
• Economics of scaling Equip. with twice
  capacity, costs less than 2 times capital
  cost is reduced /unit
• hours/labor are reduced/unit output
• Quantifying Improvement
• The Log-Log Concept (Wright-1936) How to show
  that assembly time went down airplanes,
  production went up
• Problem they evolve as y axb
• see Figs. 29.1 29.3

6
Progress (Learning) Curves

• Determining a Curve - The Recipe
• Use log x vs log y for y axb
• Use x vs log y for y aebx
• Use xn vs y (n is known) for y a bxn
• Use x vs x/y or x/y vs 1/y for y x / (a
  bx)
• The Log-Log plots can be done by hand, and are
  easy to fit
• Data Manipulation
• Garbage in becomes Bible out make sure you enter
  data correctly else errors are no longer
  detected.
• Presentation Print data, tables, plots, show
  straight lines

7
Progress (Learning) Curves

• Typical Values for Organization Progress
• (see Tables 29.3 and 29.4)
  
• Better designs
• More Efficient Factory Layouts
• Economies of Scale
• Individual Learning
• Improved Technology (from suppliers), etc, etc.
• The more that can be learned, the more will be
  learned
• f(experience, extent of mechanization) see Fig
  29.5.

8
Progress (Learning) Curves

• Percent Task Time
• Manual Machine Manufacturing Progress
  
• 25 75 90
  
• 50 50 85
• 75 25 80
• Typical Values for Learning
• Two components
• Cognitive Learning (70 curve)
• Motor Learning (90 curve)
• Hence Cognitive dominates first, and then the
  Motor becomes dominant two learning
  slopes.

9
Progress (Learning) Curves
Additional Learning Issues - Forgetting -
Breaks worker forgets but learning curve get
steeper (eventually they will
rejoin the previous rate) - Age of Operator
- After 35 learning is slow (if not required
to learn in the past) -
Nervous capabilities decline more than physical
- Good Training Programs - Reduction of Info.
Processing big improvement, compared with
speed. -Higher rates of performance go
slowly less often
10
Progress (Learning) Curves
Work Factor Learning Factors see
Table 29.6 1) Around 80 curve for cycles 1-50 2)
Around 90 for cycles 50-500 Table allows
estimate of mean time/unit for lt 500
units ESTIME WFTIME SMFACT QMULT - 1
1 Example WF of 1.5 min/unit and 100 units are
to be made, then estimated time/unit 1.51.3
1.95 For 100 units 1001.95min/unit 195
min. Allowances of 15 195/0.85 229
min. It allows to estimate specific unit at
which WF 100 After 200 - 400 cycles, a WF
standard gt 90
11
Progress (Learning) Curves
Applications Cost Allocation Need to know
your costs for bid contracts Shared costs,
increased sales, improved production, costs
of all final products Caveat Low costs do not
guarantee success if prod. is
obsolete (life cycle, birth, growth, decay,
death) Scheduling Main issue Labor hr/unit As
time passes Labor should and output
Also Lot size and poor scheduling are
important.
12
Progress (Learning) Curves
Evaluation of Alternatives Great when labor
hr/unit cte. For accurate predictions, we need
a changing cost/unit New methods might be costly
dissappointment Example Manufacturing
rather than on site
construction for big clients. More
? Acceptable Days Work
see Table 29.7 Experienced Worker Units
produced/std time New worker after 2 weeks it
can be predicted if he will -No time Std is
accurate unless it is based on a run
quantity -Major cause of inaccurate time Std. is
to forget learning
13
Progress (Learning) Curves
Incentives Problems Pay by results, failure to
recognize that labor time changes with output. -
Dissatisfaction easy jobs vs tough jobs (less vs
more) - Output restriction -
Workers Dont report output increase
complaints high wages from fellow workers.
- Employees increase their leisure
time.