Design for eXcellence

Unit 3: Design for assembly

Section 1: Design for Assembly

This unit is divided into three parts. The first describes the common DfA methods for mechanical assemblies and the second and third deal with electronic DfA.

The unit begins with an introduction to DfA and the difference between mechanical and PCA DfA. The types of DfA analysis are discussed and Function Analysis is mentioned as an example of how other techniques can be adapted to improve the DfA analysis process.

Mechanical DfA is discussed first, where some sample guidelines are provided and the three most popular DfA methods, Lucas DFA, Hitachi’s AEM and BDI’s DFA, are described in detail.

The DfA of Printed Circuit Assemblies is discussed next. Some of the popular methods used in the electronics industry are described, and we then reapply DfA principles as they apply to the electronics assembly industry.

Specific DfA guidelines for assemblies with SMT, THT components and mixed technology are described, and following these some other general PCA guidelines are listed.


Section Contents


Introduction

Design for Assembly (DfA) is one of the many philosophies and methodologies used to reduce product cost and time-to-market and improve product quality. DfA accomplishes this by providing design advice on how the product can be most efficiently and economically assembled. This advice consists of good design practice rules and guidelines for use by product designers. They are based on how the decisions made during design affect assembly. This same philosophy applies, for example, to Design for Test (DfT) and the design of the testing procedure of a product.

This philosophy, encompassed in the term Design for Excellence (DfX), is vital because 75–85% of the cost of a product is determined during its design, including component material and manufacturing choices and assembly process decisions.

By applying DfA analysis to their designs, many leading companies such as Ford, Kodak, General Motors, IBM, NCR, Xerox and more have saved millions. Cost reductions of 20–35% are commonly achieved through the use of the DfA methodology. DfA can easily achieve substantial reductions in assembly costs.

Most designers are aware of DfT, DfA, DfF (Design for Fabrication) and their meaning – DfA is covered in detail in this Unit while the other two are presented separately. We can, however, take one step back, and look at the wider context of ‘DfM’, which to some is an umbrella term embracing DfT, DfA and DfF, and to others is the process of analysing the board for any PCB fabrication, assembly or test issues before design sign-off.

In some companies, different groups of engineers are responsible for DfT, DfA and DfF. Other companies use a single DfM engineer to cover all three. The latter situation would tend to favour integration of the three aspects, and this philosophy is supported in these units. The reason makes sense. The manufacturing process of a printed circuit assembly is so complicated that one aspect of the design may affect its fabrication, assembly and testing. When analysing a design, therefore, all three must be considered. For example, pad spacing rules relate to solder bridging in assembly, feature resolution in fabrication and probe clearances in test. In practice, a DfM engineer who has compiled a checklist (or set of checklists) for the organisation’s designs will, over time, have incorporated design rules and constraints from all three anyway.

We can also take two steps back, to the wider context of DfX. Along with DfT, DfA and DfF, DfX can incorporate design guidelines from environmental legislation, product servicing and repair requirements, to give just three examples. Guidelines generated from any downstream activity can be included in a DfX engineer’s checklist.

Difference between mechanical and electronic assembly DfA

In most situations, PCA layout designers will be involved in the analysis of electronic assemblies. Some electronic assemblers also offer product case or housing assembly and PCA development engineers may be asked to take part in mechanical assemblies DfA. It is important to understand and appreciate these techniques and the DfA of mechanical and electronic assemblies is covered in this unit.

This definition of DfA is taken from Syan and Swift:

  1. Reduce the number of components in the product

  2. Optimise the ‘assemblability’ of the remaining parts

  3. Optimise the ‘handlability’ of the remaining parts

  4. Improve quality, increase efficiency and reduce assembly costs.

There are fundamental differences between mechanical and electronic assemblies with respect to DfA analysis. These differences dictate the method used for DfA.

Types of DfA analysis

There are generally three different methods of applying some type of a DfA process to a product. The first method was that originally used, and is the following of a general set of rules or guidelines. These rules generally are not quantitative in nature and require a human to interpret and apply to each specific and unique case. Whilst this is much better than just blindly starting each design from scratch, it does require some skill and knowledge on the part of the designer to correctly interpret and apply the rules.

The second method, devised by Boothroyd and Dewhurst, employs a quantitative evaluation of the design. Each part of the design is rated with a numeric value depending on its assemblability. The numbers are summed for the entire design and the resulting value is used as a guide to the overall quality of the design. The product is then redesigned, using the numeric values as a goal to be minimised. By concentrating on areas of the design that contribute heavily to the overall score, the effects of the redesign can be maximised. However, this again requires insight and knowledge on the part of the designer. Similar methods were put forward by Lucas and Hitachi; these, along with Boothroyd-Dewhurst’s method, are described later.

The third, and most recent, development in this area is in the automation of the entire process. By using a computer, quantitative analysis can be applied to the design. Then by building an expert system in the computer employing the general design rules, a system can be developed that can first analyse a design and then optimise it. Optimisation is achieved by repeatedly applying the rules and guidelines and evaluating the quality of each iteration.

Since the early implementations of DfA tools, steps have been taken to provide a more integrated approach. Boothroyd-Dewhurst have developed a number of Windows-based tools including DfA, but also covering a greater proportion of the product life-cycle through Design for Manufacture modules (machining, die casting, injection moulding, sheet metal working, and powder metal parts), and more recently Design for Service and Design for the Environment.

Figure 1: Screenshots of the Boothroyd Dewhurst DFMA software tool

Screenshots of the Boothroyd Dewhurst DFMA software tool

The Lucas DFA was the responsibility of Lucas Engineering and Systems. In 1995 they were taken over by CSC, a large IT service company. The take-over has seen the incorporation of the Lucas DFA into a product called TeamSET. TeamSET is an integrated suite of formal methods implemented around a common relational database. The tools available in the package include, DfA, manufacturability analysis, FMEA, concept convergence, QFD, and design to target cost.

Figure 2: Screenshot of the TeamSET DFA software tool

Screenshot of the TeamSET DFA software tool

AT&T has an extensive computer analysis program that analyses circuit board designs for assemblability that began in 1985. Since that time, an extensive array of rules (comprising over 30 categories) has been compiled. A computer program has been designed to allow the design engineer to immediately see the application of the rules to his particular design. Beginning with the selection of the actual components, the system is used to automatically populate the board according to the design rules.

Specialised components may be placed by hand as needed. Throughout the entire process (component selection and placement, interconnection routing, testability, and manufacturability), the software continually monitors the design, flagging any areas that violate design rules. Since the beginning of the DfM program, the time for the design and analysis of a circuit board has dropped from several weeks to 30-45 minutes.

At the moment, the most comprehensive DfX software tool (including DfA) commercially available for PCAs is offered by Valor Computerised Systems. Its origins were in the optimisation of bare board fabrication processes. Valor expanded the tool to include component assembly on bare boards and the provision of assembly line optimisation and computer-aided manufacture programme generation.

Figure 3: Screenshot from the Valor Trilogy tool

Screenshot from the Valor Trilogy tool

This shows the main graphical manipulation toolbars at the top, colour-coded layer information at the left-hand side. and a main screen showing the layer details.

Adoption of other tools into DfA

DfA is only one of the tools available to the design engineer producing a design (or redesign) and requiring recommendations for optimisation. In this unit Functional Analysis is mentioned because it provides the analysis team with another tool to optimise product cost. It is useful because it questions the need for a particular component to be included in an assembly in order to meet the product’s functional specification, whereas DfA assumes that all the functions of the product are required and provides recommendations for easier assembly. It is, however, a waste of time and money to assemble components that are not required by the customer. For example, components are sometimes added to a product as a quick fix to a problem: subsequent redesigns solve the problem, but, because the engineer who added the quick fix has retired, nobody knows why it was added! The assumption is that every component has a critical function that should not be removed.

We can claim the same advantages from any of the cost reduction and optimisation tools available, but there are potentially huge rewards from Functional Analysis, and it would be relatively simple to incorporate it in a DfA process.

Functional Analysis

Functional Analysis is a very useful and popular technique that can be used alone or as part of another system. It forms the basis for the Value Methods (Value Analysis, Value Engineering and many more), Quality Functional Deployment and FMEA (Failure Mode and Effects Analysis).

The process of challenging the existence of each component in a product is key to efficient assembly. Products that consist of the minimum number of parts are not only enhanced for assembly but also provide knock-on benefits through reduced stock holding and inventory, reduced manufacturing or sourcing costs, and increased reliability.

Functional Analysis is similar to DfA because they are both methods of reducing the number of components in an assembly and then highlighting the remaining components that add unnecessary costs. There are two differences between the two methods. Firstly, when Functional Analysis asks the question “Is this part necessary?” it considers whether the functional specification requires the function of that component. Therefore, instead of considering the part’s materials and relative motion as in DfA, it questions the necessity of its function. Secondly, functional analysis highlights those components whose cost is high in proportion to the importance of its function to the product.

For example, consider a suitcase that costs £25 to manufacture, with a wheel runner assembly that costs £20. Functional analysis will identify the wheel runner assembly as too expensive for the function it carries out on the product and highlight this for redesign. In contrast, DfA will only highlight whether or not it is easy to assembly. Ideally, Functional Analysis should be carried out before DfA is used. It will eliminate components that are not required to satisfy the functional specification and highlight any components that are too expensive for the function they perform. These components will form part of the redesign.

There are two advantages to analysing the function of a component:

Function Cost Matrix

The Function Cost Matrix allows an at-a-glance summary of a product’s functions. These give a pictorial view of the functions and how each of the components contributes to their costs. An example of a torch is shown in Table 1. It is constructed by listing the components of the torch in the far left column and the functions along the top row. The matrix is assembled by allocating a portion of the component’s total cost to each of the functions of that component. This requires some experience, guesswork and approximations, but once this is done the function columns can be summed to give the cost of each of the functions. It is then possible to compare each function and component cost to the others to find the expensive components or functions.

Table 1: Function Cost Matrix example – Torch
Class of function secondary secondary secondary primary secondary primary    
Function / Component protect bulb retain spring permit access provide light hold lens power bulb component cost total % of total assembly cost
lens
£0.10
£0.10
13.2
rear cap
£0.02
£0.02
2.6
threads on rear cap and body
£0.05
£0.05
6.6
bulb
£0.23
£0.23
30.3
front cap
£0.06
£0.06
7.9
threads on front cap and body
£0.05
£0.05
6.6
battery
£0.25
£0.25
32.9
                 
total function cost
£0.10
£0.02
£0.10
£0.23
£0.06
£0.25
£0.76
% of total assembly cost
13.2
2.6
13.2
30.3
7.9
32.9

Mechanical DfA

Photo 1: Examples of the type of mechanical assemblies and cases that come with PCAs

Examples of the type of mechanical assemblies and cases that come with PCAs

Mechanical DfA guidelines

The original philosophy of DfA, that is to reduce component count and ensure assemblability and handlability of the remaining components, has been built upon and expanded by many sources. Here are some generic guidelines devised by Leaney and Wittenberg in 1992:

  1. Reduce the part count and types

  2. Modularise the design

  3. Strive to eliminate adjustments

  4. Design parts for ease of feeding or handling

  5. Design parts to be self aligning and self locating

  6. Ensure adequate access and unrestricted vision

  7. Design parts so they cannot be installed incorrectly

  8. Use efficient fastening or fixing techniques

  9. Minimise handling and reorientation

  10. Utilise gravity

  11. Maximise part symmetry.

Also Corbett, Dooner, Meleba and Pym (1991) suggested a checklist for DfA should include the following:

  1. Minimise

    • Parts and fixings

    • Design variants

    • Assembly movements

    • Assembly directions


  2. Provide

    • Suitable lead in chamfers (radii on corners)

    • Natural alignment

    • Easy access for locating surfaces

    • Symmetrical parts or exaggerated symmetry

    • Simple handling and transportation


  3. Avoid

    • Visual obstructions

    • Simultaneous fitting operations

    • Parts which tangle or nest

    • Adjustments that affect prior adjustments

    • The possibility of assembly errors

DfA specific rules

At this point it is worth looking at more specific component redesign examples. Lots of great examples and more information on DfA rules can be found at Hugh Jack’s Claymore site at the School of Engineering at Grand Valley State University. The link is http://claymore.engineer.gvsu.edu/eod/design/design-52.html. It has plenty of useful examples of:

Some examples of the kind of issues described on the site are shown below.

Parts reduction

Figure 4: If screws must be used, try integrating washers with the screw heads; this will eliminate at least one part

If screws must be used, try integrating washers with the screw heads; this will eliminate at least one part

Figure 5: Replace separate springs with parts with thin sections that act as springs

Replace separate springs with parts with thin sections that act as springs

Part symmetry

Figure 6: Improving handlability of a part by redesign

Improving handlability of a part by redesign

Figure 7: Example of using symmetry to improve part installation

Example of using symmetry to improve part installation

[back to top]


DfA quantitative evaluation techniques

Three of the better-known quantitative evaluation techniques are those of Boothroyd-Dewhurst (USA), Lucas (UK) and Hitachi (Japan). All three have seen use in industry. The Hitachi Assembly Evaluation Method (AEM) was first developed in the late 1970s, with Design for Assembly (DFA) being introduced later (around 1980) to reflect the work of Professor Geoffrey Boothroyd at the University of Massachusetts. Professor Boothroyd and Dr Peter Dewhurst then used the term Design for Manufacture and Assembly (DFMA) to reflect their continued work into manufacturability analysis (machining, die-casting, powder metal, etc.) and assembly techniques.

These techniques are evaluative methods that rate or score the assemblability of designs at an early stage in the design process. They use their own synthetic data to provide guidelines and metrics to improve the design in its ability to be assembled. All three can analyse a product for manual assembly, and Lucas and Boothroyd-Dewhurst can also analyse automated assembly. DFMA goes one step further by considering different types of automated assembly such as robotic and high speed dedicated. The techniques all work in a slightly different manner in order to achieve the evaluation.

Self Assessment Questions

The three DfA analysis methods described in this unit are Boothroyd-Dewhurst’s DFA/DFMA, Hitachi’s AEM and Lucas’s DFA. Each of these uses the same basic principles of Design for Assembly for assemblability analysis. One difference between them is the way they quantitatively rate or score each design.

Can you find two ways of assessing or benchmarking a product for assemblability during Design for Assembly analysis?

There are numerous rating factors used for scoring or rating assemblability. Given the DfA philosophy of reducing the number of components then improving each component’s assemblability, can you list some of the indicators of assemblability?

Compare your list with the description we give in the next section.

Assessing a design for DfA

DfA scoring or rating methods are used by each DfA technique for assessing a product design according to its overall compliance to the requirements of the assembly line. Although the three methods use the same DfA philosophy they differ in the way they quantitatively rate or score each design.

In general, there are two ways. One compares the number of parts in an assembly and how each part is assembled with a perceived ideal. This ideal will have the least number of parts possible and each part will be optimised for assemblability. After DfA analysis, redesign recommendations will take two things into account: the number of parts to be as close the ideal number as possible and second, product and assembly process assembly is as close to the ideal as possible. The other way uses a rating system for comparing number of parts and assemblability for alternative designs. The rating system can be assembly cost, effort or time based. Scores will be based on the number of parts and how easy it is to assemble those parts. The best design is the product with the highest (or lowest) score.

The way each technique quantifies the assemblability is the biggest difference. By definition, a quantitative technique uses numerical scores, but how do you score whether a component is necessary or not? To reduce the number of components, Boothroyd-Dewhurst’s DFA advises the analysis team to ask a series of questions for each component:

  1. During operation of the product, does the part move relative to all other parts already assembled?

  2. Must the part be of a different material from all other parts already assembled? Or isolated from them?

  3. Must the part be separate from all those already assembled because otherwise necessary assembly or disassembly of other separate parts would be impossible?

If the answer to all of these questions is no, then that part is unnecessary and the analysis recommends redesigning the product so that the function of the part is carried out by another component. The component is then scored 0. If the component was necessary, that is the answer to any of the above questions was yes, then the component would score 1. The Boothroyd-Dewhurst metric NM – the Theoretical Minimum Number of Parts – is the sum of the scores. In Boothroyd-Dewhurst’s case this is used to derive an Ideal Assembly Time, which can be compared to the actual assembly time.
Other factors are assembly direction – where the ideal is vertically down using gravity; handling and feeding; and assembly cost. Each technique uses its own factors and also the metrics based on those factors. These will be covered in more detail in the next section where full descriptions of the three methods are given.

Hitachi Assembly Evaluation Method

Assembly Evaluation Method, AEM, was developed by Hitachi Corp. Tokyo, Japan. Its main objective is to facilitate design improvements by identifying ‘weaknesses’ in the design at the earliest possible stage in the design process, by the use of two indices:

Assemblability evaluation score ratio (E), used to assess design quality by determining the difficulty of operations

Assembly cost ratio (K) used to project elements of assembly cost. The procedure of the analysis in the AEM method is as follows (Figure 8):

Figure 8: Hitachi’s AEM procedure

Hitachi’s AEM procedure

The total assemblability evaluation score for the product is defined as the sum of the assemblability scores for the individual tasks, divided by the number of tasks. This may be considered to be a measure of design efficiency where a score of 100 would represent a perfect design. Hitachi consider that an overall score E of 80 is acceptable and overall assembly cost ratio K of 0.7 is unacceptable.

Redesign of a simple product using AEM

An illustration of a simple redesign procedure is shown in Figure 9, Figure 10 and Figure 11. It shows the original and two redesigns of a block attachment to a chassis.

Step 1: (Original design)

Here it is necessary to attach a small block B, to a chassis A. The initial method, shown in Figure 9, involves the use of bolt C.

Figure 9: Original design

Original design

 

Table 2: Evaluation score and the cost ratio of the original design
  part assemblability evaluation score E
assemblability evaluation score
K
assembly cost ratio
Set chassis A
100
73
1
Bring down B and hold it to maintain orientation
50
fasten screw C
65

Step 2: (Redesign 1)

Examining the original design, the holding down to maintain orientation is the worst individual evaluation score and the suggestion is that the need for holding be removed by spot-facing the chassis shown in Figure 10. This gives an improved evaluation score and cost ratio as a result of this (Table 3).

Figure 10: Redesign 1

Redesign 1

 

Table 3: Evaluation score and the cost ratio of redesign 1
  part assemblability evaluation score E
assemblability evaluation score
K
assembly cost ratio
Set chassis A
100
88
0.8
Bring down B (orientation is maintained by
spot-facing)
100
fasten screw C
65

Step 3: (Redesign 2)

Here, the bolt has been removed and the block attached to the chassis by using a press fit. The assembly evaluation score for the press fit is less than that for simple block placement and reduces from 100 to 80 but, importantly, one part has been eliminated. As a result, although the product evaluation score has not significantly improved (89 compared with 88), the assembly cost ratio has significantly improved because of the reduced number of parts (Figure 11 and Table 4).

Figure 11: Redesign 2

Redesign 2

Table 4: Evaluation score and the cost ratio of redesign 2
  part assemblability evaluation score E
assemblability evaluation score
K
assembly cost ratio
Set chassis A
100
89
0.5
Bring down and pressfit block B
80
 

Lucas DFA method

The Lucas method is a result of the collaboration between Lucas and the University of Hull. It does not use cost analysis, and in this respect differs from the Hitachi and Boothroyd-Dewhurst methods. The method involves the assigning and summing of penalty factors associated with potential design problems, similar to the Hitachi method but with the inclusion of handling (or feeding) as well as insertion. These are denoted in a visual flow called an Assembly Flow Flowchart (ASF). Figure 12 and Figure 13 show the ASF for the drain pump example as a series of symbols representing a gripping process, insertion process, etc. Three scores, design efficiency, feeding/handling ratio and fitting ratio, are generated in three stages of analysis:

These scores can then be compared to thresholds or to values established for previous designs.

Redesign of a drain pump assembly using Lucas DFA

Figure 12 and Figure 13 show how the Lucas method can show design efficiency and the feeding/handling and fitting ratios and how redesigning a product can improve them.

Design efficiency

The first Drain Pump design shows a poor design efficiency of 4 essential parts out of 25 (16%). The second design reduces the number of parts to 6 with a resulting design efficiency of 66%.

Feeding/handling and fitting ratios

A visual comparison between the two ASFs show that the redesign is much simpler to assemble. This is reflected in the difference between the two Fitting Ratios (4.0 for the redesign compared to 19.9 for the original).

The redesign has reduced the number of components by getting rid of the bolts, washers and nuts. These happen to attribute the higher feeding analysis scores (and fitting) to the total (this makes sense because these tend to be difficult to assemble in reality). As a consequence the feeding and handling ratio has reduced from 6.9 to 1.63.

Figure 12: Original drain pump assembly design

Original drain pump assembly design

 

Figure 13: Redesign using the Lucas DFA method

Redesign using the Lucas DFA method

Boothroyd-Dewhurst DFA

The Boothroyd-Dewhurst DfA evaluation centres on establishing the cost of handling and inserting component parts. The process can be applied to manual or automated assembly, which is further subdivided into high-speed dedicated or robotic. An aid to the selection of the assembly system is also provided by a simple analysis of the expected production volume, payback period required, number of parts in the assembly, and number of product styles.

Regardless of the assembly system, parts in the assembly are evaluated in terms of ease of handling and ease of insertion, and a decision is made as to the necessity of the part in question. The findings are then compared to synthetic data, and from this a time and cost is generated for the assembly of that part. The opportunity for reducing this is found by examining each part in turn and identifying whether each exists as a separate part for fundamental reasons. These fundamental reasons are:

  1. During operation of the product, does the part move relative to all other parts already assembled?

  2. Must the part be of a different material from all other parts already assembled? Or isolated from them?

  3. Must the part be separate from all those already assembled because otherwise necessary assembly or disassembly of other separate parts would be impossible?

The second stage of the analysis is to examine the handling and insertion of each component part. For manual assembly, a two digit handling code and a two-digit insertion code are identified from synthetic data tables. The tables categorise components with respect to their features for handling such as size, weight, and required amount of orientation. For insertion, there are categories for part alignment, the type of securing method, and whether the part is secured on insertion or as a separate process. These codes are then cross-referenced to identify the time for that operation from the table.

The codes and subsequent times are used to determine a number of metrics:

Though costs and times are determined, care must be taken in the use of these values in an absolute sense. As with other techniques, values are best used for comparing redesigns.

Figure 14: A flow diagram of the Boothroyd-Dewhurst DFA method

A flow diagram of the Boothroyd-Dewhurst DFA method

Redesign of a simple product using the Boothroyd-Dewhurst DFA method

Figure 15: Original design

Original design

Figure 15 shows a simple sub-assembly used in the construction of a gas-flow meter. The objective is to analyse the design using the Boothroyd-Dewhurst method with the intention of using the information obtained to create a new, easier-to-assemble, less expensive sub-assembly. In this analysis only manual assembly will be considered. For the redesign of an existing product, it will be assumed that the functional parts must have the same dimensions and be made of the same materials.

Table 5: Manual assembly worksheet for the original design
part no number repeats total assembly time manual assembly cost min number parts remarks
6

2

6
1.2
0
nut
5
2
6
1.2
0
washer
4
1
4
0.8
1
plate
3
1
3
0.6
0
bearing housing
2
2
20
4
0
screw
1
-
-
-
-
complete assembly
39
7.8
1
design efficiency = 3 * min parts / assembly time = 3 * 1 / 39 = 0.077

Table 5 shows a design for a manual assembly worksheet for the product shown in Figure 15.

If at least two parts are necessary, these would have to be the bearing housing and the plate, as these are both functional, and all the other parts merely fasteners. There are many possibilities for joining the bearing housing to the plate using integral fastening: one proposed solution is by the use of integral rivets as shown in Figure 16. The worksheet for this solution is shown in Table 6.

Figure 16: Redesign 1

Redesign 1

 

Table 6: Manual assembly worksheet for Redesign 1
part no number repeats total assembly time manual assembly cost min number parts remarks
3
1
3
0.6
1
bearing housing
2
1
4
0.8
0
plate
1
-
-
-
-
complete assembly
7
1.4
1
design efficiency = 3 * min parts / assembly time = 3 * 1 / 7 = 0.428

The plate can be placed either way up, but it does have rotational asymmetry and is ‘thin’. One solution is to have one axially-symmetric integral fastener as shown in Figure 17. For this solution, the bearing housing cannot be improved since it still needs to be assembled one way up, and one of two ways round, but the plate is now easier to handle. The worksheet for this solution is shown in Table 7.

Figure 17: Redesign 2

Redesign 2

 

Table 7: Manual Assembly Worksheet for Redesign 2
part no number repeats total assembly time manual assembly cost min number parts remarks
3
1
3
0.6
1
bearing housing
2
1
3
0.6
0
plate
1
-
-
-
-
complete assembly
 
6
1.2
1
 
design efficiency = 3 * min parts / assembly time = 3 * 1 / 6 = 0.5

 

Requirements of a PCA DfA process

Imagine that you have been requested to design a DfA process for your organisation. The obvious starting point is to define a list of requirements for the process. Can you list the main requirements?

Compare your answer with this one.

[back to top]