Taguchi’s methods incorporate the use of statistical techniques. Taguchi defines the quality of a product as the loss imparted by the product to the society from the time’ the product is shipped. Taguchi's methods emerged because of his disagreement about the use of Zero Defect as a principle to produce quality products.
Details about Taguchi:
- Taguchi’s methods incorporate the use of statistical techniques.
- They are primarily intended for designers and engineers to optimize the settings so that products are robust. These statistical methods are intended as a troubleshooting/problem solving tool in the early stages of the product development cycle.
- Besides control variables which are dealt with by SPC, Taguchi methods enable engineer designers to identify ‘noise variables’ which if not controlled can affect product manufacture and performance.
- Taguchi defines the quality of a product as the loss imparted by the product to the society from the time’ the product is shipped.
- The loss may include various things such as customer complaints, added warranty costs, damage to company reputation, loss of market lead amongst others.
L = D2C(Loss increases by the square of deviation from the target value)
- Taguchi argues that a product does not start causing losses until it is out of specification but more importantly when there is deviation from the target value.
- Factory B runs higher risk loss costs (quality loss factory loss). The more deviation there is from targets, the greater are the losses.
- The Quality Loss Function (QLF) can also be represented by a quadratic formula
QLF is useful because it not only warns about deviations at the early stages of new product development but can also provide managers with cost estimates.
Zero Defect principle:
- Taguchi's methods emerged because of his disagreement about the use of Zero Defect as a principle to produce quality products.
- The Zero Defect principle is that robustness derives from consistency.
- Provided that there is a consistency in deviations, it will be quite possible to make adjustments in the target. Zero Defects does not permit scattered deviations within specifications.
Taguchi's quality imperatives:
1. Quality losses result from product failure after sale; product 'robustness' is more a function of product design than on-line control, however stringent, of manufacturing processes.
2. Robust products deliver a strong 'signal' regardless of external noise and with a minimum of internal 'noise'.
3. To build robust products, set ideal target values for components and then minimise the average of the square of deviations for combined components, averaged over the various customer-user conditions.
4. Virtually nothing is gained in shipping a product that just barely satisfies the corporate standard over a product that just fails. Get on target; don’t just try to stay in spec.
5. A concerted effort to reduce product failure in the field will simultaneously reduce the number of defectives in the factory. Strive to reduce variances in the components of the product and variances will be reduced in the production system as a whole. Competing proposals for capital equipment or competing proposals for on-line interventions may be compared by adding the cost of each proposal to the average quality loss, that is, the deviations expected from it.