Standards (Contd.)

Too many operations-based organisations delude themselves on standards. It is tempting to simply impose work standards by getting industrial engineers or work study people to do the work and to post standards at workstations, or worse to keep them in a file. Then of course, there is no buy-in, no foundation for continuous improvement, and worst of all a great likelihood of large process variance. So the quick way turns out to be the least effective way and often a total waste. Traditional "Taylorism", is suitable only the sweat shop. The Toyota way takes time to train operators to do the analysis and standardisation themselves - they need to not only learn basic work-study principles but also to appreciate the reasons why they should do it. Toyota regards work standardisation as one of its most challenging management tasks. Robert Hall believes that this worker-oriented standardisation, requiring high levels of operator skill and motivation, is one of the main limitations to the growth of Toyota.

So what should a standard contain? Operator standards should contain the takt time, the detailed work sequence steps that are involved together with the time taken for each step (these should be written in the operators own words), and the standard inventory quantity or kanban quantity involved during the takt time cycle. These should be supported by a diagram of the operator movements, possibly a sketch of any operation requiring clarification, any tools that are used, any quality checks, and a note on any special operation that is required to be carried out periodically. Standard procedures can be colour coded to match the product that carries a label of matching colour. When an engineering change occurs, a number on the product should match the number on the standard sheet. Often, a standard should cover not only what to do when things are normal, but also what to do if things go wrong. Consider health and safety. There must be own responsibility for the standard, for publishing it, for keeping it at the workplace, and for keeping it up to date. At Intel where consistency is an absolute requirement, operators audit one another using the standard instructions.

When cellular manufacturing is established or when takt times change, operators in the cell are encouraged to rebalance the line themselves or to change their own work standards to the new takt time. In this case there may be work standard for the whole cell, including a diagram showing the movements of each operator.

Managers and other workers should also consider documenting their standards. At least there should be broad standards for repetitive activities such as meetings, the budget cycle, personnel procedures, document preparation, placing of orders, appraisals, and so forth. The broad steps or a loose descriptive format may be appropriate. The same PDCA thoughts apply; if new form of meeting is found to work better than the old one, document both with comments on why the new works better. But do it, otherwise the opportunity to learn may be lost.

Robert Hall has pointed out that there are three stages of standardisation. The first is "outcome only" where the standard is established in a plan or drawing but how to get there is not specified. This generally leads to a high level of variation, and the need for inspectors. The second stage is "standardised processes" where variance is reduced by following standardised methods to achieve the outcome, supported in good cases by for instance SPC and pokayoke. The third stage is "standardised predictive methods" whereby a high level of consistency is achieved directly from the design by using standardised processes that work first time "without the need to prototype, test, or make special tools". Very few companies or processes are at this stage.

Further reading
Robert W Hall, "Standard Work: Holding the Gains", Target, Fourth Quarter, 1998, pp 13- 19
Taiichi Ohno, Toyota Production System, Productivity Press, Portland, OR, 1988
Edward E Lawler III, From the Ground Up: Six Principles for Building the New Logic Organisation, Jossey-Bass, San Francisco, 1996, Chapter 6