Overlapping sequential design activities based on concurrent engineering a Technique

Overlapping sequential design activities is a strategy developed based on concurrent engineering principles that allows reducing the time usually required to complete project design. Reducing design delivery time allows construction to start sooner, thus leading to reduction of overall project delivery time. One way to reduce overall project delivery time is by adopting concurrent, overlapped design processes by overlapping dependent activities instead of following traditional sequential processes.

Following concurrent engineering practices, overlapping strategies resort to reducing or

removing information dependencies among activities by altering their existing characteristics to create a more favorable environment for activity overlapping. The extent to which two activities can be effectively overlapped depends on the relationship between them. Prasad identifies 4 possible types of relationships between activities (Prasad 1996): 1) dependent activities, 2) semi-independent activities, 3) independent activities, and 4) interdependent activities. When two activities are dependent, the downstream activity requires information from the upstream task before the downstream task can begin. Semi-independent activities require only partial information from the upstream activity before the downstream activity can be started. Independent activities require no information from one activity before the other activity can begin. Interdependent activities require a two-way information exchange between them before either can be completed (Prasad 1996).

Independent activities can be overlapped without any risk of delay or rework because the upstream activity does not require information from the downstream activity to begin. Dependent activities, on the other hand, carry risk when overlapped. When overlapping dependent activities, the downstream activity begins before all the information from the upstream activity is available, thus, the downstream activity begins with incomplete, non- optimal, or non-final information (Bogus et al. 2005). Changes in the upstream activity can also impact the downstream task, resulting in potential delays and/or rework. Because of this risk involved in overlapping dependent activities, this technique focuses on developing strategies to reduce the dependencies between these.

The degree to which dependent activities can be overlapped is determined by the nature of information exchange between them. The information exchange between an upstream activity and a downstream activity can be described in terms of the natural rate of information evolution in each activity and the sensitivity of the downstream activity to changes in upstream

information (Bogus et al. 2005). Thus, activities’ characteristics of evolution and sensitivity are used to determine appropriate strategies for achieving overlap to reduce design delivery cycles.

Information evolution

The natural evolution characteristics of an activity determine the rate at which information is generated when no time constraints or pressures are applied. In a traditional design process, work is performed following the natural evolution characteristics of activities; therefore, activities are carried out only when all upstream information is available. However, traditional approaches do not always allow for the most effective design delivery process in terms of time. Hence, evolution characteristics can be used in project scheduling decisions to identify potential opportunities for overlap to reduce design time.

According to Bogus et al.’s research work, there are four essential determinants of an activity’s evolution (Bogus et al. 2005):

o Design optimization o Constraint satisfaction

o External information exchange o Standardization

Design optimization refers to the level of optimization achieved by design elements or the number of design alternatives evaluated. For example activities that require the evaluation of many alternatives will have slower evolutions than those that require only one or a few alternatives. Constraint satisfaction refers to the flexibility of design elements in satisfying constraints such as physical limitations. External information exchange refers to the amount of information received from or reviewed by external sources. Activities that require information from external sources may result in multiple iterations of design. These activities will have a slower evolution than activities that do not require external information exchanges.

Standardization describes the level of standardization in the design product and/or the design process. Standardization allows activities to have faster evolutions (Bogus et al. 2005).

Information Sensitivity

Sensitivity refers to the amount of rework that a downstream activity will have to go through if information on the upstream activity changes. So, a highly sensitive activity will require a larger amount of rework if upstream information changes even when the change is minimal.

Bogus et al. define the following as the main determinants of sensitivity in design activities (Bogus et al. 2005):

o Constraint sensitive o Input sensitive o Integration sensitive

Activity sensitivity can be determined by the proximity of the downstream design to boundaries or constraints. When a downstream design element is near a certain type of constraint, such as a maximum or minimum capacity performance, the changes in upstream information can lead to significant rework in the downstream activity. Input sensitive refers to the level of dependence of downstream tasks on specific inputs from other activities. Integration sensitive involves the ability of downstream design elements to be separated from the entire system (Bogus et al. 2005).

b. Implementation

Based on activities’ characteristics of evolution and sensitivity, Krishnan et al. define overlapping in four possible situations (Krishnan et al. 1995). The first one is given when evolution of the upstream task is fast and the sensitivity of the downstream task is low. This is the most convenient situation for overlapping, which is highly recommended through exchange of preliminary design information and early finalization of the upstream design. This strategy is termed distributive overlapping. The second situation is given when activity evolution and sensitivity are both low. Under these circumstances overlapping is recommended only through

the exchange of preliminary design information, called iterative overlapping. When evolution is fast but sensitivity high, only early finalization of upstream information is recommended, referred to as preemptive overlapping. Finally, when evolution of the upstream activity is low and sensitivity of the downstream task is high, overlapping should occur to the least degree possible. In this situation, the strategy recommended is to decompose activities into sub- activities or packages known as divisive overlapping (Krishnan et al. 1995).

Upstream activities with fast evolution and downstream activities with low sensitivity represent the better combination for effective overlapping. Thus, overlapping strategies should aim at changing the evolution upstream activities from its natural state to a faster state to speed up the design process of the activity. Sensitivity characteristics are more likely to be affected by the design situation; in consequence, unlike evolution there are no natural characteristics that determine the sensitivity of downstream tasks to upstream information changes. Once activities characteristics of evolution and sensitivity have being defined and characterized, adequate overlapping strategies can be applied to speed up activities evolution and reduce activities sensitivity. Bogus et al. suggest the following (Bogus et al. 2005):

Strategies that speed up evolution

Early freezing of design criteria

Early freezing of design criteria consists on releasing information from an upstream activity to the downstream activity before the upstream design is complete. This strategy requires project participant’s commitment early in the design process to generate the specific required design criteria as soon as possible.

Advantages

By early freezing design criteria, some of the uncertainty on downstream design is reduced by eliminating the likelihood of changes in upstream information when downstream activities have already begun.

Disadvantages

Early freezing design can lead to increased project costs due to lack of design optimization. There is also risk that the pre-established criteria may not be feasible in all situations, therefore, the risk of rework in downstream activities that have already begun based on the initial design criteria increases.

Design quality and final products may also be affected by the loss of information quality in the upstream activity at the time of freezing.

Key elements to ensure a high degree of success

Early freezing of design criteria is recommended only when the upstream activity is fast evolving.

Early release of preliminary information

Early release of preliminary information from the upstream activity also enables the downstream activity to begin before the upstream activity is completed.

Advantages

This strategy allows downstream activities to begin faster than traditionally based on preliminary information, resulting in potential reduction of overall design delivery time. Disadvantages

The risk associated with the early release of preliminary information is that this information might change as the upstream activity is finalized. If changes happen, the downstream activity may require rework, resulting in extra costs and delays. The impact that changes in the upstream design have on downstream activities is directly related to the amount of overlapping between activities; thus, the more the overlap, the greater the impact that upstream design changes have on downstream activities, and the higher the amount of rework.

Key elements to ensure a high degree of success

This strategy is only recommended when the downstream activity has low sensitivity to changes in upstream information.

Prototyping

Prototyping is the process of quickly compiling preliminary upstream design information into a working model of the ultimate system (Bogus et al. 2005). This model is a preliminary prototype which serves as a basis for discussion and revision among project designers to produce the final product based on the preliminary prototype.

Advantages

Prototype models allow the downstream activity to proceed when the working model is finished, before the actual upstream activity is finalized. Prototyping is very suitable for complex

systems, where there are many pieces of information to pass to downstream activities. Disadvantages

Prototyping is based on early criteria which typically require substantial revisions before the activity can be completed. This introduces a high risk of significant rework on downstream activities and the related costs and delays.

Key elements to ensure a high degree of success

This strategy is only recommended when the downstream activity has low sensitivity to changes in upstream information.

No iteration or optimization

This strategy is applied to activities with a naturally slow evolution, where iteration or

optimization delays the availability of upstream information to downstream activities. Thus, this overlapping technique recurs to placing time constraints by limiting the number of iterations allowed in an upstream activity before passing the information for downstream design. Advantages

Limiting iteration or optimization speeds up the evolution of a slow evolving activity, which allows information to be passed to downstream tasks earlier. Through this strategy, downstream activities start faster to reduce project design cycles.

Disadvantages

The lack of design optimization introduces substantial risks of significant rework on the downstream activity leading to increases in project costs and potential delays.

Key elements to ensure a high degree of success

By definition, this strategy is only applicable to upstream activities with a slow evolution. Standardization

Standardization refers to the adoption of design practices to be used repetitively on a project (Gibb 2001). This technique aims at the adoption of standardized products, components or designs to accelerate the natural evolution of an upstream activity so information can be released to the downstream activity earlier.

Advantages

Standardization expedites the transmission of upstream information to the downstream task, reducing design delivery time. If properly applied, standardization may also decrease project costs by eliminating sub-optimal designs (designs in which only one designer has optimized its part) and by increasing constructability (Bogus et al. 2005).

Disadvantages

Similar to other strategies, standardization involves the risk of project cost increases and delays due to possible rework as a result of lack of design optimization.

Key elements to ensure a high degree of success

Again, this strategy is only applicable to upstream activities with a slow evolution by definition, as fast evolving activities are already standardized.

Strategies that reduce sensitivity

Overdesign

Overdesign relies on the adoption of conservative assumptions. By making conservative assumptions, it is possible to work in the downstream activity before the upstream activity is completed, and in some cases, before the upstream activity has even begun.

Advantages

Starting downstream activities before the upstream activity is completed or has begun allows starting activities faster, reducing thus overall design delivery time.

Disadvantages

The risk involved in overdesigning is that the assumptions made might not be conservative enough leading thus into having to redesign and rework on the downstream activity. Therefore, overdesigning presents the risk of increasing project costs and delays because of rework. Key elements to ensure a high degree of success

The extent to which sensitivity is reduced depends on the quality of information used for overdesign in the downstream activity. In consequence, faster evolving activities are, by nature, more likely to provide better information to develop overdesign assumptions for downstream design. Nonetheless, overdesign is also recommended to upstream activities with slow

evolution as this strategy based on conservative assumptions for design. However, more risk is taken when applying overdesigning strategies with slow evolving upstream activities.

Set-based design

This technique refers to the parallel development of multiple upstream designs to decrease the sensitivity of downstream activities. In a set-based design, a designer develops a set of solutions for one component in parallel with designers of other components. As design progresses, the set of solutions are gradually narrowed. However, designers agree to stay within a narrowed

predetermined group of solutions; therefore the final design represents a final integrated solution of the individual designs that falls into the pre-established solution set.

Advantages

Set-based design allows designers to develop downstream design sets at the same time that upstream activities are designing their sets, which reduces the sensitivity of downstream

activities to changes in upstream activities. Design delivery time is reduced because downstream design is developed earlier in the process.

Disadvantages

This strategy presents a major disadvantage. Developing multiple designs for each activity or a more conservative single design increases design costs.

Key elements to ensure a high degree of success

Set-based design is best applicable when the upstream activity has a slow evolution as the strategy assumes the development of multiple alternative upstream designs.

Decomposition

This strategy consists on the decomposition of one activity into smaller packages of activities with faster evolution characteristics. Decomposition can also be applied to downstream activities to reduce their sensitivity. The objective of decomposition is to create new activities that can be overlapped using any of the previously mentioned overlapping strategies.

Advantages

Decomposing one activity in smaller packages create new upstream activities with faster

evolution and new downstream activities with lower sensitivity which create better opportunities for overlapping by reducing the risk of rework and delay.

Disadvantages

Decomposition involves re-analysis and double overlapping work of the new activities created. Key elements to ensure a high degree of success

This strategy is only recommended when no other overlapping strategy is effective as it involves double overlapping work.

Enhanced overlapping strategy framework

Choosing the most appropriate strategy depends on the evolution and sensitivity characteristics of design activities but also the specific project conditions. Aligning strategies at the determinant level provides information about which of the strategies are most appropriate for a given

context. Figure 3 is a basic framework that presents the appropriate strategies to be applied depending on the evolution and sensitivity characteristics of a pair of dependent activities (Bogus et al 2005).

Evolution

Slow Fast

Overdesign Early release of preliminary info Early freezing of design

Prototyping Overdesign No iteration/optimization Early release of preliminary info

Standardization Prototyping Set-based design Low Overdesign

No iteration/optimization Early freezing of design

Standardization Overdesign Set-based design Decomposition Sensitiv ity High

Figure 3. Basic overlapping strategy framework (taken from Bogus et al. 2005, pp. 19)

c.

Advantages

The major advantage brought by this technique is the potential reductions in design delivery. Overlapping sequential dependent design activities allows reducing the time normally required to complete project design, which therefore allows earlier design releases for construction execution. By adopting concurrent and overlapped design processes, construction can be expedited resulting in overall project schedule acceleration.

d. Key elements to ensure a high degree of success

Overlapping strategies always involve some level of risk. Therefore, its implementation has to follow a systematic analysis and thorough process to identify which overlapping strategies should be adopted and when these should be implemented to minimize the risks of delays and rework. The project or construction manager also needs to be aware and prepared to quickly respond if the overlapping strategy fails and take effective actions to mitigate any problems.

e.

Disadvantages

In general, overlapping strategies for reducing project delivery time involve a certain amount of risk. The risk depends on the assumptions in which the decisions of beginning the downstream activity are based and the sensitivity of the downstream activity to changes on the assumptions made when all upstream information is finalized.

As discussed before, activities with low sensitivity can be overlapped with lower risk of delay and rework than activities with high sensitivity. However activities with low sensitivity are not completely free of risk. It is not always possible to speed up the evolution characteristic of an activity as desired, and starting a low sensitive activity before all upstream information is complete also involves a certain degree of risk of delay and rework if the upstream information changes.

The most common dependencies among design activities are information and resources. The overlapping techniques presented focuses on information dependencies and the sensitivity of these dependencies to changes in upstream information. Thus the technique assumes that there are enough resources available to eliminate resource dependencies between activities, which is seldom the case in real life.

Other risks and costs associated with overlapping include lack of design optimization and coordination, increased materials wastage, frequent change orders, inadequate coordination between design and construction, and inadequate scheduling of the work package interfaces. Other consequences include increased costs because of increased coordination loads in terms of the volume and frequency of communication between project team members. And, as

concurrency increases on a project, the coordination requirements also increase along with its related coordination costs (Fazio et al. 1988, William 1995).

f.

Applicability and use

There are a few steps that can be followed to enhance overlapping strategies applicability. The first step is to develop a critical path network schedule for the design process without considering any overlap between activities. The critical path schedule provides a basis against which time savings from overlapping strategies can be measured.

Next, activities that belong to the critical path should be identified, along with the evolution and sensitivity characteristics of each one. This step can be accomplished by using the key

determinants of evolution and sensitivity suggested earlier. Time savings are achieved only when activities on the critical path are overlapped.

Once the evolution and sensitivity characteristics of dependent activities on the critical path have