LEARNING
CURVES
Dr. Anurag Tiwari
IIM Rohtak
LEARNING CURVE
CONCEPTS
A learning curve is a line displacing the relationship between unit production time and the
cumulative number of units produced.
Predicts reduction in manufacturing costs or direct labor hours as cumulative production
increases.
Based on empirical evidence rather than theory.
Learning Curves can be applied to individuals or organization .
Individual learning is improvement that results when people repeat a process and gains skill
and efficiency.
Organizational Learning results from practice as well, but it also came from changes in the
administration, equipment, and product design.
ASSUMPTION IN LEARNING
CURVE
Learning curve theory is based on three assumptions:
The amount of time required to complete a given task or unit of a product will be less each time
the task is undertaken.
The unit time will decrease at a decreasing rate.
The reduction in time will follow a predictive patterns.
In Manufacturing Case:
Process should be repetitive
Process should more labor intensive
Least employee turnover
LEARNING CURVES AS TIMES
AND NUMBER OF UNITS
Cumulative Average Output Per
Time per Time Period
Unit
Average
Observed Output during
Data Fitted Time
a Time Period
in future
Unit Number Time
A Progress Curve Industrial Learning
FACTORS AFFECTING THE
LEARNING CURVE
Worker’s Work
Material
Skill Method
Product Learning
Tools
Design Curve
Continuous Process
Improvement Design
MODELS
Cumulative average time learning models (Wright Model)
Incremental unit time learning model (Crawford Model)
CUMULATIVE AVERAGE TIME
LEARNING MODEL
Learning increase with a fixed percentage every time when production become double
Formula
Total Time= Initial Time*(2*LC)*(2*LC)*(2*LC)…
WRIGHT LEARNING CURVE
Production
Cost
Unit number
WRIGHT LEARNING CURVE
In Wright's Model, the learning curve function is defined as follows
Y = aXb
where:
Y = the cumulative average time (or cost) per unit.
X = Unit number
a = time (or cost) required to produce the first unit.
b = slope of the function or learning decimal percentage.
= log of the learning rate/log of 2.
LEARNING CURVE-80%
Unit numbers Unit Direct Cumulative Cumulative Avg
Labor Cost Direct labor Direct labor
Hours Hours
1 100 100 100
2 80 180 90
4 64 314.20 78.553
8 51.2 534.592 66.824
Can calculate with log formula-
= 100(8)^log0.8/log2
EXAMPLE
Suppose the first 20 units take 20 hours to be produced . The learning curve rate is 90%.
40 Units=20(2*0.9)=36 hours
80 Units=20(2*0.9)(2*0.9)=64.8 hours
160 Units=20(2*0.9)(2*0.9)(2*0.9)=116.64 hours
320 Units=Units=20(2*0.9)(2*0.9)(2*0.9)(2*0.9)=209.95 hours
Total time taken by all units-Time taken by the previous unit
2nd lot of 20 units took (36-20)=16/20=0.8 hours/unit
EXAMPLE
Aerosub Inc. has developed a new product for spacecraft that includes the manufacturing of a
complex part. The manufacturing of this part requires a high degree of technical skills.
Management believes there is a good opportunity for its technical force to learn and improve
as they become accustomed to the production process. The production of first unit requires
10000 direct labor hours. If an 80% learning curve is used and eight unit are produced, the
cumulative average direct labor hours required per unit of the product will be:
A. 5,120 hours
B. 6,400 hours
C. 8,000 hours
D. 10,000 hours
EXAMPLE
Unit 1= 10,000 hrs.
Unit 2= (10,000*(2*0.8))=16,000
Unit 4= (10,000*(2*0.8)(2*0.8))=25,600
Unit 8= (10000* (2*0.8)(2*0.8)(2*0.8))=40,960
Cumulative Direct labor hours= 40,960/8=5,120 hrs.
INCREMENTAL UNIT TIME
LEARNING MODEL
Every incremental unit takes less time at a fixed percentage due to the effect of learning curve.
Assumption:
Time declined by a constant percentage each time the cumulative quantity doubles. (i.e. 1 st, 2nd,
4th, 8th, 16th,… unit)
Formula
Total Time= Initial Time*LC*LC*LC…
EXAMPLE
Suppose the first units take 20 hours to be produced . The learning curve rate is 90%.
2nd unit= 20*0.9=18 hous
4th unit= (20*0.9)*0.9=(18*0.9)= 14.58 hours
8th unit =14.58*0.9=13.12 hours
16th unit=13.12*0.9=11.81 hours
EXAMPLE
Propeller Inc. plans to manufacturer a newly designed high-technology propeller for airplanes. Propeller
forecasts that as worker gain experience, they will need less time to complete the job. Based on prior
experience, propeller estimates a 70% cumulative learning curve and has projected the following costs.
Cumulative number of Manufacturing Projections
Units produced Average Cost per unit Total Cost
1 20,000 20,000
2 14,000 28,000
If Propeller manufacturer eight propellers, the total manufacturing cost would be a
A. $56,660
B. $54880
C. $62,643
D. $112,000
EXAMPLE
Unit 2 Cost=14,000
Unit 4th cost = (14000*0.7)= 9800
Unit 8th cost =(9800*0.7)=6860 cost per unit.
Total Cost= 8*6860=54,880
ESTIMATING THE
LEARNING PERCENTAGE
The following guidelines are useful for estimating the impact of learning on manufacturing
tasks.
These guidelines use estimates of the percentage of time spent on manual work versus the time
spent on machine-controlled work.
75 percent hand assembly /25 percent machine = 80 percentage Learning
50 percent hand assembly /50 percent machine = 85 percentage Learning
25 percent hand assembly /75 percent machine = 90 percentage Learning
SET OF GUIDELINE IN
SPECIFIC INDUSTRY
Aerospace, 85 Percent
Shipbuilding, 80-85 Percent
Complex machine tools for new model, 75-85 percent
Repetitive electronics manufacturing, 90-95 percent
Repetitive machine or punch-press operations, 90-95 percent
Repetitive electrical operation(wiring and circuit board fabrication), 75-85 percent
Repetitive welding operations, 90 percent
Raw material manufacturing, 93-96 percent
Purchased parts fabrication, 85-88 percent
How Long Does Learning Go On ?
INDIVIDUAL LEARNING
Employee Applicant A Employee Applicant B
2.0 2.0
1.5 1.5
Time
(minutes) 1.0
1.0
0.5 0.5
2 4 6 8 10
2 4 6 8 10
Repetition number Repetition number
INDIVIDUAL LEARNING
General Guideline to improve individual performance based on learning curves includes the following
Proper Selection of Workers: A test should be administered to help choose the workers. These tests should be
representative of the planned work.
Proper Training: The more effective the training, the faster the learning rate.
Motivation: Productivity gains based on learning curves are not achieved unless there is a reward.
Work Specialization: As a general rule, the simpler the task, the faster the learning.
Do one or very few jobs at a time: Learning is faster on each job if completed one at a time, rather than working on
all jobs simultaneously.
Use tools or equipment that assists or support performance:
Provide Quick and easy access for help: The benefit from training are realized and continue when assistance is
available.
Allow workers to help redesign their tasks: Taking more performance factors into the scope of learning curve, in
effect, shift the curve downwards.
ORGANIZATIONAL LEARNING
LEARNING CURVE
APPLICATIONS
Production planning / EOQ planning
Price forecasting
Petrochemicals
Consumer durable goods
Competitive bidding
Income reporting in accounting
Planning warranty maintenance
Washers / dryers
Televisions
Forecasting industrial accidents
Petroleum industry
Mining
Forecasting automobile accidents on new roadways
LEARNING CURVE
EXAMPLES
TABLE E.1 Examples of Learning-Curve Effects
LEARNING-
IMPROVING CUMULATIVE CURVE
EXAMPLE PARAMETERS PARAMETER SLOPE (%)
1. Model -T Ford Price Units produced 86
production
2. Aircraft Direct labor-hours per Units produced 80
assembly unit
3. Equipment Average time to replace a Number of 76
maintenance at group of parts replacements
GE
LEARNING CURVE
EXAMPLES
TABLE E.1 Examples of Learning-Curve Effects
LEARNING-
IMPROVING CUMULATIVE CURVE
EXAMPLE PARAMETERS PARAMETER SLOPE (%)
4. Steel Production worker Units produced 79
production labor-hours per unit
produced
5. Integrated Average price per unit Units produced 72a
circuits
6. Handheld Average factory selling Units produced 74
calculator price
LEARNING CURVE
EXAMPLES
TABLE E.1 Examples of Learning-Curve Effects
LEARNING-
IMPROVING CUMULATIVE CURVE
EXAMPLE PARAMETERS PARAMETER SLOPE (%)
7. Disk memory Average price per bit Number of bits 76
drives
8. Heart 1-year death rates Transplants 79
transplants completed
9. Caesarean Average operation time Number of 93
section baby surgeries
deliveries