|
|
www.johnstark.com |
|
|
|
A variety of parameters such as product/market features, activities, costs, time, resources, number and frequency of iterations, number and frequency of changes, and extent of parallel activities characterize engineering workflow. The activities carried out by a particular company will depend greatly on the type of product and the company's position in the value chain. Within process industries, there will not be the same workflow for a producer of fine chemicals as for a producer of bulk chemicals. Similarly, in discrete manufacturing industries, the workflow will be different for producers of airplanes, machine tools and durable consumer goods. Even between companies in the same sector, workflow will be different. Differences will be due to different customer bases, product life cycles, production runs, new product introduction rates, and special regulations, as well as to company size and organizational structure. Workflow is company-specific. There is no easy solution available allowing a company to improve its workflow without first understanding all the details of its workflow. The approach to changes, of both product and process specifications, also differs from one company to another. Changes cost money, waste time, require iterations in the workflow, and require extra management effort. When design changes occur late in the workflow, they can be extremely expensive, impacting the product, production equipment and the launch date. In the worst case, the customer will receive a defective product. It is important to understand where a company defines costs in the workflow, and where costs are actually incurred. Various studies have been made of the distribution of costs in the traditional engineering workflow. Typically they show some 60-75% of costs defined (and 1-5% of costs incurred) during conceptual design, and 85-95% of costs defined (and 10-15% incurred) before release to manufacture. The incurred cost of engineering changes during conceptual design is negligible compared to that of changes after release. Fixing a design problem when a product is in the field is often thousands of times more expensive than preventing it during initial design. It is better to spend more effort on getting the conceptual design right, than on doing it quickly and making expensive corrections later. The above figures do not take account of product use by the customer. Some products require a great deal of support and maintenance. The price of maintenance and spare parts for some products exceeds their purchase price. For example, airlines typically spend 2 or 3 times the purchase price of an aircraft engine on spare parts for the engine. Support and maintenance activities can significantly change the distribution of life cycle costs, often reducing even further the percentage of costs incurred during the early, development phases. They should be considered as part of the engineering workflow, and taken into account when deciding how to improve the workflow, and where to assign resources. An inefficient workflow The traditional engineering workflow is inefficient. Unnecessary steps, errors and changes make it expensive, slow, bureaucratic, and, invariably, longer than expected. It suffers from low quality, poor communications, a lack of management understanding, and a lack of structure. Unlike the manufacturing process, for which the workflow is defined in detail, the workflow in the engineering process is generally not well structured, except in those areas where it is bureaucratic. There is no real Quality Assurance or Quality Control in the engineering workflow - or so it would appear from the number of engineering changes that are due to errors, not enhancements. Often there are no formal procedures for estimating the engineering time and cost. In the absence of any clear structure, it is difficult to optimize the workflow, with the result that estimates of, for example, lead times, will invariably be incorrect and longer than necessary. The lack of workflow structure makes it difficult to manage the time and cost of a given project, with the result that engineering projects are often late and over budget. In many cases, the skill and experience of the engineering managers involved reduce these problems. However, the opposite is generally true in cases where management does not have a good understanding of the workflow. More overhead is then added, in the form of unnecessary management reports, which are sent to people who neither understand their contents nor are capable of acting on them. Serious communication problems may occur at the borders between functional organizations. Design engineers pass manufacturing engineers designs that can not be produced. The design has to go back for correction. Engineering changes costing thousands of dollars result. The engineering/marketing, engineering/finance and engineering/field frontiers are also potential sources of problems. Sales people offer customized versions without knowing if it will be possible to make them profitably. Design engineers are unable to get cost information from the finance function. Design engineers do not receive field reports, and design existing problems into new products. Maintenance requirements are not taken into account during conceptual design. The traditional engineering workflow does not encourage users in different functions to communicate freely and prevent these problems. |