لطفا منتظر باشید ...
13.3 The Process Recombinator
The Process Recombinator is a software tool that uses the Process Handbook to support a process innovation methodology as described above. This methodology consists of three key steps (Herman et al. 1998 contains a detailed explanation of the methodology):Identify the core activities and the key dependencies (i.e., the deep structure)ofthe process you want to redesign, using the process specialization hierarchy.
Systematically generate a set of alternative refinements (i.e., surface structures)for the tasks and dependencies in this deep structure model, by ''recombining''existing or newly generated alternatives for these process components.
Select from this set the process(es) that appear to best satisfy your requirements, possibly using information stored in trade-off matrices.
We will describe how these steps are accomplished in the s below. The capabilities underlying steps 1 and 3 are part of the original set of Process Handbook tools, so we will summarize them quickly and then focus on the Recombinator capabilities of step 2.To illustrate the capabilities of the Recombinator tool, we will use examples based on a field study we conducted in collaboration with one of our corporate research sponsors, the AT Kearney consulting firm, and one of their clients which we call Firm A to preserve the client's anonymity (for more detailed descriptions of this study, see Malone et al. 1999; Herman et al. 1998; Kruschwitz and Roth 1999).Firm A was experiencing increasing problems with their hiring process. They were growing rapidly in a tightening labor market, and they had a culture of independent, competitive business units. Together, these factors led to increases in the time and cost to hire people and to increasingly frequent instances of business units ''hoarding''candidates or bidding against each other for the same candidate. In an effort to improve their hiring process, the organization had invested a great deal of time and energy into ''as is''process analysis using conventional techniques such as flow-charting. But they also wanted some way to come up with highly innovative ideas about how to improve their process.The Recombinator was completed after the field study. The examples shown here demonstrate how the tool now supports the manual process that was followed in the field study.
13.3.1 Identifying the Process Deep Structure
The first step in our methodology is to identify the deep structure, that is, a process model that captures the essence (the core activities and key dependencies) of the process we wish to redesign. This maximizes room for new ideas by abstracting away nonessential features. The Handbook supports this via the specialization hierarchy. Users can either select an existing generic process from the hierarchy or create a new one. Since Firm A wanted to improve their hiring process, we use the ''Hire''process as the starting point for our example scenario (see figure 13.5).
Figure 13.5: The deep structure for 'Hire'
13.3.2 Process Recombination
The next step is to find alternative ways (i.e., different surface structures) for implementing the generic activities and coordination mechanisms identified in the deep structure model. This is achieved by the Process Recombinator, which includes three parts. The three parts can be used independently; each allows systematic exploration along a different set of process design dimensions. First, we will look at the sub-activity recombinator, which generates all possible combinations of the specializations of the subactivities in the process. Next, we will consider the dependency recombinator, which generates different combinations of coordination mechanisms for the process dependencies. Finally, we will look at the bundle recombinator, which generates different combinations of the alternatives in the dimensions represented as bundles.This order of usage was chosen for illustrative purposes only, however. The three parts of the Recombinator can be used in different sequences depending on one's needs.
The Subactivity Recombinator The subactivity recombinator lets users pick different specializations for each of the subactivities in a process (see figure 13.6). For example, the ''Select human resources''subactivity of the hiring process has specializations such as (1) 'Select by role-playing'(e.g., a process used by Cessna to screen candidates for executive positions), (2) 'Select based on education'(a screening process implicitly used by many management consulting firms), and (3) 'Select by attrition'(a screening process used by universities who admit all applicants and then fail many of them in the first year). Using the Process Handbook capabilities, users can easily see more detailed descriptions (and other information) about each of these activities.
Figure 13.6: Subactivity recombinator user interface
As the figure shows, each subactivity is placed in a separate column, and each column contains the alternative specializations for that subactivity. Using this display, users select the combination of specializations they want to use in creating a new process. The system then automatically generates the new process specified. If users make multiple selections in some of the columns, then all combinations are generated. Figure 13.7 gives an example of the process created for the selections made in figure 13.16. (Users who want to know more about how the alternatives in a given column compare can click on the ''trade-off''button for the column and see a trade-off matrix for those alternatives.)
Figure 13.7: Results of using the subactivity recombinator
The power of this approach is that the specialization hierarchy allows the process designer to draw on relevant ideas and insights from many different kinds of organizations, opening the possibility of useful new combinations never before considered in a particular setting.
The Dependency Recombinator The dependency recombinator complements the subactivity recombinator by allowing one to also consider alternative coordination mechanisms for process dependencies. Instead of displaying only the subactivities of a process, it displays both the subactivities and dependencies as a flowchart (figure 13.8).
Every subactivity and dependency can have an associated list of alternative choices below it. The lists below dependencies allow users to select the coordination mechanisms used to manage them. In figure 13.8, for example, we can see different alternatives for managing the dependency between ''Use headhunter for sourcing''and ''Select human resources . . .''There could be a traditional ''push-based''coordination, where the headhunter contacts the firm. Alternatively there could be an ''open market''of 'sellers'(headhunters and internal HR departments) and buyers (line-function departments). The ''market with bonus''coordination mechanisms reimburses the seller (in our case the headhunter) with a fee depending on the new employee's performance in the firm. This encourages headhunters to think about the long-term performance of a candidate. Once the user has selected alternatives for each subactivity and dependency, the system automatically generates new process designs in the same way as the subactivity recombinator.
Figure 13.8: Dependency recombinator user interface
In bringing in coordination possibilities from far afield (e.g., as on-line bidding systems for internal recruiting), this approach can generate very innovative process possibilities.The Bundle Recombinator The bundle recombinator helps users generate new design alternatives by exploring the multiple possibilities defined by bundles in the specialization hierarchy. Consider, for example, the specialization subtree under 'Install Employee'(figure 13.9).[1]
Figure 13.9: Specialization sub-tree for 'Install employee'
Recall that the bundles under a given process in the specialization hierarchy group together refinements of that process that differ along a particular dimension such as who does the work and how it is done. Generally, each bundle captures an orthogonal design dimension. The four bundles under 'Install employee'therefore define a four-dimensional space of possible combinations (e.g., figure 13.10 shows the combinations defined by two of these dimensions).
Figure 13.10: Part of the design space for the 'Install employee'process (the cell marked is the example described in the text)
Each cell in this four-dimensional space represents a possible new process specialization formed by making one selection from each bundle dimension. For instance, figure 13.11 shows the combination 'Install by oneself during work within the job environment'. One example idea stimulated by this combination is training novice air traffc control offcers by interleaving simulations of unusual situations in the middle of their real work environment (perhaps without the trainees even knowing that these were simulations).
Figure 13.11: Bundle recombinator user interface
Another interesting combination (stimulated by the combination of two dimensions marked by an asterisk in figure 13.10) is to let new employees ''install''themselves by having them decide what they could do best for the firm. In this process new employees look around to find something useful to do for the firm (Kaftan and Barnes 1991 report this type of behavior in some of the hires at SUN Hydraulics).
When users have selected a combination of alternatives (e.g., as in figure 13.11), they press the button shown in the upper left corner of the figure, and the system generates the new process they have specified. The subactivities in the newly created process are derived by ''multiple inheritance''from the ''parent''processes (Wyner and Lee 1995). The algorithm used for multiple inheritance is as follows: Sub-activities that appear in one or both specialization parents are inherited as is. If one parent process has a more specialized form of a subactivity than the other does, then the more specialized version of that subactivity is inherited. If one parent process has deleted a subactivity that appears in another, or if a subactivity is specialized in different ways in the different parent specializations, then the system asks the user what to do. The power of this approach is that it can lead to novel combinations that enable process designers to be more creative in their process designs.The Roles of the Different Recombinators As we have seen, the subactivity and dependency recombinators have similar functionality while the bundle recombinator takes an orthogonal approach. All three approaches can be used in a fully integrated way. The subactivity recombinator is useful when we wish to focus on alternatives for the core activities in the process. The dependency recombinator is useful when we wish to also explore different ways of managing the key dependencies in the process.
Figure 13.12: Trade-off matrix for new process re-designs. (All values are for illustration purposes only.)
The bundle recombinator, finally, allows us to create new process specializations suggested by using bundles as design dimensions. The new specializations created by any Recombinator tool can then, of course, be used as alternatives within the other ones. The decision of which Recombinator to use first is dependent on what aspect of a process seems to be most promising for generating novel processes. Exploring the design space of process ideas becomes an iterative process in which the three parts of the Recombinator are used in turn, until a satisfactory set of interesting alternatives is generated.
13.3.3 Comparing the New Process Designs
Once users have used the different components of the Process Recombinator to produce a number of candidate process re-designs, they can use a trade-off matrix to help assess each re-design from the perspective of the criteria that are meaningful to them. The selections made in figure 13.8, for example, would yield the rows shown in the trade-off matrix in figure 13.12.
The Handbook specialization hierarchy can include, for each process, attributes and associated values that describe the process. Attributes that are potentially appropriate for comparing the newly generated process alternatives are thus automatically inherited by the new combinations just as subactivities are. Thus the columns of this trade-off matrix are also automatically generated by the system. In some cases default values for the cells will be inherited as well. It is up to the user, however, to determine whether the values of these attributes are appropriate for each alternative and to change them if necessary. Once this is done, the new processes and associated trade-off values are maintained in the Handbook repository as a source of ideas for future users.[1]The term ''install''may seem mechanistic when applied to employees. The term itself, however, suggests potentially innovative analogies with situations where other kinds of things are ''installed.''
