Total Productivity Maintenance and Learning Curve

No Comments

Print Friendly, PDF & Email


Total Productive Maintenance


Taguchi Method and Learning Curve




Total Productive Maintenance (TPM) is known as productive maintenance in Japanese industry, broadly considered an extension to Total Quality Management (TQM). Accordingly, TPM aims at keeping the current plant and equipment at its highest productive level through cooperation of all areas of the organisation. Though the aim of TPM is towards less defects, TPM also requires historical and current data and relevant statistical analysis to determine any downtime of a or all equipment and regular /periodic activities related to maintenance of equipment to ensure that they operate efficiently in the organisation. The overall goals of TPM are:

  • To maintain and improve equipment capacity
  • To maintain equipment for life
  • To use support from all areas of operation
  • To encourage input from all employees
  • To use teams for continuous improvement

Based on the above mentioned objectives, TPM can be explained as follows:

Total = All encompassing by maintenance and production individuals working together

Productive = Production of goods and services that meet or exceed customers’ expectations

Maintenance = Keeping equipment and plant in as good as or better than the original conditions at all times.

Calculation of TPM

TPM is a simple calculation of Overall Equipment Effectives (OEE) wherein,

OEE = Availability (A) * Performance (E) * Quality (R)

Availability refers to a machine or a business unit available for production when scheduled. This implies that when a process (in a machine or a business unit) is running then it creates value for the organisation and if the process is stopped then the machine or the business unit is creating costs to the organisation with no associated value. The process may be stopped either due to mechanical failures in the machines, unavailability of resources (raw materials, human resources, etc), social/political instabilities, etc. A process when stopped is broadly an indication of lost production or output (or value) due to downtime. Availability (A) can thus be calculated as follows:

Availability (A) = [Operating time (T)/Planned operating time (P)]*100, where operating time is the difference between planned operating time (P) and downtime (D). In the example discussed in class, A = [3000 (P) – 637 (D)] / 3000 = 0.7876 or 78.76%

Performance (E) generally focuses on determining any wastes created during operations, which could have been less than optimal speed. The wastes are identified by comparing the actual or current cycles to the ideal cycles times. Accordingly, Performance (E) can be calculated as follows: {[Theoretical cycle (C) * Processed Amount in quantity (N)] / Operating Time (T)}*100. In the example discussed in class, E = [0.5 (C) * 4450 (N)] / 2363 (T) = 0.9415 or 94.15%

Quality focuses on identifying time wasted by producing a product that does not meet quality standards. This calculation also attempts to determine the percentages of processed products to the rejected parts at a given time period. Correspondingly, rate of quality  (R) can be calculated as follows:

Rate of Quality (R) = {[Processed amount in quantity (N) – Non-conformities (Q)]/ Processed amount in quantity (N)} *100. In the example discussed in class, R = {[4450 (N) – 15 (Q)]/4450 (N)} *100 = 0.9966 or 99.66%

Finally, OEE equals 0.7876*0.9415*0.9966 = 0.7390 0r 73.90%. OEE is a simple metric that indicates the current status of a manufacturing process. It is a tool that combines multiple manufacturing aspects and attempts to highlight underlying problems that require improvements. In summary, OEE quantifies how well the manufacturing unit or equipment performs relative to its designed capacity at time period when it is scheduled to operate. Accordingly, OEE represents the percentage of scheduled time the operation is available to operate (Availability), the speed at which the unit or equipment runs as a percentage of its designed speed (Performance) and the percentage of total units produced without defects (Quality)



The learning curve is based on the principle that all jobs in an organisation are performed more efficiently as greater experience is gained in respective jobs. The learning curve is a mathematical tool that shows the relationship between the number of units produced with the labour hours required to produce them. Based on this relationship an organisation can predict future costs of production. The learning curve theory has been useful in industries like airframe aerospace, electronics, appliances, etc[1]. The learning curve theory can be also used for bringing about improvements in performance skills, design, planning, layout, management controls, computer programming, etc.

Learning curve was originally developed in 1936 by T. P. Wright and is a graphically representation of changing rate of learning related to time or labour hours, costs or performance. The learning curve can be measured using the following formula:

Y = aXb where,

Y = the cumulative average time (or cost) per unit required for production

X = the cumulative number of units produced

a = time (or cost) required to produce the first unit

b = the slope of the function or the learning rate/factor, which is calculated as log of learning rate/log of 2

As seen above the learning curve can be measured in terms of time and cost. In terms of time, the learning curve reflects the cumulative time saved when producing certain number of units in accordance to the learning factor/ rate. In terms of costs, the learning curve reflects the cumulative costs saved when producing certain number of units in accordance to the learning factor/rate. The learning curve picture in the powerpoint slides titled “lecture 11” is a steep (convex) learning curve indicating that the product is easy to learn. The opposite or a concave curve means that the learning curve is shallow and indicates that the learning is gradual or slow.

The learning factor/rate represents the rate of improvement in performing a task as a functionof time and a lower learning rate implies the learning is faster. For example, if the rate of improvement of 20% each time the quantity produced is doubled then the learning rate/factor is 100-20, i.e. 80%. The 80% further implies that to produce the 2nd unit of the product, it will take only 80% of the time of the first unit and the 3rd unit will take 80% of the 2nd unit and so on. For example, an employee produced 1 unit 100 minutes. Assuming that the employee’s learning rate is 80%, the employee will spends 80% less time or 80 minutes (0.8 *100 minutes) to produce the 2nd unit, and 64 minutes (0.8 * 80) to produce the 3rd unit and 51.2 minutes (0.8 * 64) to produce the 4th unit, etc. This example can also be used to understand learning curve with regards to saving costs. Assuming costs of producing 1 unit is Rs. 100 and with a learning rate of 80%, the 2nd unit will be produced at Rs. 80, the 3rd unit will be Rs 64 and so on. However, the costs reduced in terms of the learning rate should be critically evaluated in terms of inflation and should be accordingly adjusted. Cost reductions through learning curve could be considered as a bargaining tool between suppliers and the buyers and to predict future costs.

Applications of learning curve

Learning curve can be used across various activities within an organisation and some of those activities are highlighted below:

  • Labour efficiency: Employees become physically dexterous, mentally confident and spend less time experimenting and making more mistakes. They learn to take short-cuts and strive towards getting used to the processes in an organisation
  • Standardisation and Specialisation: With standardised processes, employees gain more experience with limited set of tasks and operate at a faster rate. Accordingly, employees also specialise in limited and standardised set of tasks thus gaining specialised experience
  • Technology-driven learning: Learning curve theories encourage technologically innovative and automatic equipment and processes further driving efficiency and effectiveness within an organisation
  • Better use of equipment: The application of learning curve encourages productive use of equipment led by increased production and lower costs of production
  • Changes in resource mix: As company acquires experience, resource mix can be altered or modified leading to efficient use of resources and reducing wastes
  • Product redesign: With manufacturers and consumers having more knowledge and experience about the product, the product can be redesigned or improved. For example, mobile phones can be operated by simply touching the screen in comparison to using the keypad in the old phones.
  • Network building and use-cost reduction: When the product is used by all, everyone can use the product efficiently with learned familiarity. For example, social networking site like Facebook can be used to promote your business
  • Shared experience effects: Any efficiency learned from one product can be applied to other products. This can be experienced when two or more products share a common activity or resource. For example, riding a bicycle and a two-wheeler or using a remote for TV and DVD players, etc

[1] Purchasing and Materials management, By Gopalkrishnan