Value-added analysis of the construction submittal process

Automation in Construction 20 (2011) 1070–1078

Contents lists available at ScienceDirect

Automation in Construction j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / a u t c o n

Value-added analysis of the construction submittal process E. William East ⁎, Danielle R. Love Engineer Research and Development Center, United States

a r t i c l e

i n f o

Article history: Accepted 5 April 2011 Available online 25 May 2011 Keywords: Business process reengineering Total quality management Value-added analysis Construction contract administration Project management Submittals

a b s t r a c t Construction contracts require builders to reproduce and certify paper documents that attest to the quality and or functionality of the systems, equipment, products, and materials planned for inclusion in a finished facility. Once submitted these documents are approved or acknowledged as part of a quality control process prior to placing purchase or fabrication orders. Despite widespread advances in other areas of information technology, the submittal process remains mired in paper-based procedures that add unnecessary cost to construction projects. This paper describes current practice and then highlights lessons-learned from selected interviews with those adopting innovative technology to streamline that process. The authors discuss their submittal process simulation model. Construction teams may adapt this model to reflect their specific context to justify and monitor the application of new technologies aimed at eliminating submittal transmission, routing, and handling costs within the context of existing contracting processes. Published by Elsevier B.V.

1. Introduction Business process analysis is often complicated because managers frequently are not responsible for processes affecting other business units outside their span of control [10]. In large corporations, top management must create staff-level, cross-organizational teams to quantify how self-optimizing behavior of individual business units that decreases overall organizational effectiveness. Once the team is in place, there are a number of different methods used in business process studies. While these methods often evaluate a variety of metrics [20] common denominators will be familiar to construction managers — reducing cost and time. In practice, improved business process based on new technology is typically restricted to improvements of existing workflows [5] since the manager is unable to control the broader contractual and business environment beyond their span of control. The analysis of cyclic process waste was the basis for the total quality management and business process reengineering movements that spread throughout virtually all major industry segments, with the exception of construction and agriculture, in the late twentieth century. The resulting business processes, such as the logistics management prowess of Wall Mart, are widely recognized as the basis for the success of these firms. Recognition of the value of business process even went so far as to view business process as patentable intellectual property. A recent study by the National Research Council stated that one effective way to motivate improve-

⁎ Corresponding author. E-mail addresses: [email protected] (E.W. East), [email protected] (D.R. Love). 0926-5805/$ – see front matter. Published by Elsevier B.V. doi:10.1016/j.autcon.2011.04.008

ments within the construction industry might be to focus not on productivity increases but on decreasing waste [17]. All public projects, regardless of size, have a set of contract specifications that identify the quality and/or functionality of systems, equipment, products, and materials used for that project. Bidders, contractors, sub-contractors, suppliers and manufacturers use these specifications during the initiation and execution of construction contracts. Operators and maintenance staff use these specifications to determine characteristics of required replacement systems and equipment. Renovators use these specifications to determine the “typical” existing conditions within a facility they are preparing to modify. While there are a number of different ways to approach specifications, the most common types are performance-based and proprietary. Performance-based specifications stipulate a product based on performance criteria. Performance specifications typically include three commercial products meeting the required criteria or allow the contractor to substitute an equivalent product. Proprietary specifications identify individual product by specific manufacturer and model number. Proprietary specifications may also allow for an “or equal” product. Proprietary specifications lead to more contract disputes, with performance based specifications being “a more absolute and inviolable technique than trying to reference a standard by naming a competitor's model” [11]. This is because the phrase “or equal” leaves the specification open for a wider interpretation than if the product is selected based on performance. In addition to the specification of products through either performance or proprietary requirements, specifications also define the process through which evidence of specification compliance is documented. U.S. federal contracts use the Unified Facilities Guide Specification “UFGS 01-33-00 Submittal Procedures” [15], or similar, to

E.W. East, D.R. Love / Automation in Construction 20 (2011) 1070–1078

identify their requirements. UFGS 013300 states that the construction contractor is provided with a list of all submittals, produced during the design contract, called a “submittal register.” The construction contractor is required to physically provide a minimum of six (6) paper copies of each submittal for review and approval or simply for acknowledgement of receipt. UFGS 013300 has no provision for the submission of electronic submittal of documents. The submittal process is slightly different for design-bid-build contracts when compared to design-build contracts. The major difference is in the position of the designer with respect to the evaluation of submittals. In the design-bid-build framework, the designer is a reviewer of site-manufactured components and is contracted by the owner's representative. In the design-bid-build framework, the designer (called the “designer of record”) provides an internal review of documents at the request of the construction contractor. 2. Objective The objective of this research project was to determine the applicability of value-added analysis to identify waste inherent in construction administration processes. If successful, this technique may be used to evaluate the costs and benefits of technologies to reduce this waste, and by extension project overheads. The quantification of waste can translate into the business cases needed by project managers to justify the adoption of new technologies. To accomplish this objective, the authors' created a process model describing the submittal processing requirements for typical specifications of United States public agencies. The authors held a series of interviews with members of several project teams using who used different technologies when performing the required submittal processes. Three methods identified were (1) typical paper-based submittal process; (2) email distribution of e-documents, and (3) extranet exchange of e-documents. Differences in the nature of the work performed in each of these three scenarios were identified and described in this paper. 3. Previous effort To date there has been limited research aimed at evaluating contract administration tasks within the construction industry. A qualitative model of construction claims processing within construction companies, notes this lack of documented construction management process (Kululanga 2001). Efforts to develop semi-automated workflow management tools [13] reinforce the local optimization of business processes since target implementation of such tools would necessarily be within the context of a single organization. Attempts at developing discrete-event simulations for the Bidder Inquiry business process [6] have provided convincing arguments for the development of robust models demonstrating change management. The complexity of such models appears, however, to be out of reach for most practitioners. Trade publication discussing the submittal process [14] identifies only specific portions of the current submittal process without identification of a standard methodology that would assist future researchers when considering other processes. Efforts to streamline product selection and purchasing using centralized electronic catalogs have not been widely adopted. Major product publication houses have been unable to bridge the gap between the concept of electronic product data exchange and the practice of construction submittals and product selection. While some researchers have described the potential use of electronic product data to streamline procurement cycles [4], there remains a wide gap between such individual research studies and current practice [1]. Trischler introduces the “Value-added assessment” framework adopted as the basis of this research [20]. Value is defined subjectively, in the context of lean manufacturing, as those tasks

1071

that materially contribute to the quality of the project obtained by the consumer [9]. Trichler provides a catalog of value and non-value added activity classes used in the manufacturing industry [20]. The value-added assessment method begins with the decomposition of business processes into activities mapped to their responsible organizations. The graphic format of this model is called the “swimlane” diagram. To determine the level of activity detail within the model tasks that share resources, within a given firm, need not be further disaggregated [21]. Additional decompositions of such activities are only required if the responsible organizations have high internal communications costs. The contribution of the value-added approach is the explicit representation of activities that contribute to the value of a product, and those that do not directly contribute to the value of that product. Moving bricks around a construction site several times prior to their installation is an example of non-value added activity. Placing bricks into the final wall is an example of a value added activity. Few who have spent time running construction contracts in the field would argue that administrative procedures on construction contracts are replete with non-value added activities.

4. Model development Based on a review of the value-added methodology proposed by Trischler, the authors identified five class of value added and nonvalue added administrative activities types present in the submittal processing procedure. This initial model was developed based on standard specifications, authors' experience, and trade publications [3]. This model was discussed and validated during interviews with practitioners as described later in this section. In this model, there is one single type of value-added activity, the “processing” activity. Processing tasks are those that require the intellectual contribution of the assigned resource. The task of discovery of product data meeting specific contract requirements by a prime or subcontractor is an example of a processing class of activity. For the development of the model, the specific content of a given processing task is of no importance, the aggregation of all activities of that type is the goal of the model. Non-value added activities present in administrative processes can be classified by Trischler as “routing,” “handling,” “transmitting,” and “terminating” activities. “Routing” tasks are those tasks needed to determine the routing of specific information through a workflow that contains parallel paths. For example, the contractor must decide if they or the subcontractor is responsible to prepare a given submittal. A “handling” task is required to model any activities associated with tracking or distributing information within a given firm. For example logging the receipt of a document is a handling task. Since distribution of information affects the resulting business process [12], the model also includes an activity type called the “transmission” activity. Transmission tasks are highly dependent upon the software support system support context. These contexts based on the dynamic contractual relationships among business partners [18]. To complete the set of model classes needed for the submittal register process modeling a terminator entity that starts or ends the process is added. The authors applied these generic classes of value-added and nonvalue added tasks, in the construction submittal process, by creating swim lane diagrams for each of three phases of the contractually required submittal process. The first phase is the “preparation” of submittals by the construction contractor and subcontractors. The second phase is the “review” of these submittals by the owner's representative with assistance, as required, by the design team. Once the submittal has been approved, the contractor or subcontractor “initiates” the purchase orders for the products to be installed. The model assumes a design-bid-build submittal process where none of

1072

E.W. East, D.R. Love / Automation in Construction 20 (2011) 1070–1078

the companies were geographically collocated. Company personnel are assumed to be collocated. The authors' model was validated based upon interviews with the following sources: (1) a construction firm with the only previously published results on their use of electronic submittal transmission [2], (2) several architectural firms participating in the Specifiers Properties information exchange project [7], and (3) several government contract administrators employed at the National Aeronautics and Space Administration, U. S. Army Corps of Engineers, and U.S. Department of State, Overseas Buildings Operations office. Two of the interview scripts used for some of the initial interviews conducted by the second author are provided in the Appendix. Several of these interviewees are specifically identified in Acknowledgments. Durations were based upon interviews and checked against two contractual constraints on the “review” phase of the process. The maximum allowable duration for acknowledgment of contractor certified submittals is a maximum of thirty (30) calendar days. The maximum allowable duration for shop drawing review is forty-five (45) calendar days. Interviewee feedback indicated that durations measured in increments of ½ of a calendar day are satisfactory since such a unit of measure would allow the model to ignore the variable nature of administrative operations such as the cost of telephone calls, time taken for coffee breaks, and interruptions. Without a wide-ranging statistical study the expected values provided by this anecdotal evidence allows the model to be calibrated such that the results are within the range of practitioner expected results. As shown in Fig. 1, the submittal process for the typical federal design-bid-build contracts begins with the general contractor determining if they, or the subcontractor, will purchase the required submittals described in the product specification sections. Suppliers or manufacturers provide the necessary product data. According to those interviewed approximately, the majority of pre-manufactured product information is available on the World Wide Web. These documents allow the contractor to demonstrate that the selected Prime Contractor

Sub Contractor

1.01 START

1.02 Submittal Register

1.03 Routing

1.05 Log Request

1.04 Discover Product

1.06 Discover Product

4.1. Processing tasks

1.07 Package Submittal 1.09 Evaluate Compliance

1.08 Log Submittal

product(s) meet the specified technical requirements. Subcontractor documents are forward through the prime contractor for final review and certification. Typically, both the contractor and each subcontractor keep their own set of log files that typically do not link back to the underlying product data submitted or the locations of these components within the facility. In this simplified model we assume that all subcontractors follow the process of the idealized subcontractor from Fig. 1. If the subcontractor submittal is incomplete or does not meet the contract requirements, the general contractor will make the correction or return it for revision. For the purpose of our first order model, the recursive loops that would be included in a full model of the business process are not included. Once the individual submittal package components are reviewed for compliance, the contractor repackages the submittals. This repacking is required to transform submittals provided according to work-packages into the productbased submittals specified in the construction contract and listed in the submittal register. The contractor certifies the submittal by ink signature on a standard paper submittal routing form, copies the routing form for each of the six or more paper packages needed, and transmits all supporting documents to the owner's representative. Once submittals are received by the owner's representative, the second stage of the process begins, as shown in Fig. 2. The designated owner's Quality Assurance Representative (QAR) reviews the submittal to determine if the products listed satisfy the design requirements. Depending on the nature of the submittal and the expertise available at the architecture firm, the QAR may do this check himself or send the submittal by postal mail to the owner, tenant representatives, or the architect for consultation. If the architect or consultant determines that the submittal does not satisfy the design requirements, the submittal will be marked as “rejected” or “resubmit as noted” and mailed to the QAR who logs the reply. Next, the QAR transmits the reply back to the contractor by postal mail. The general contractor mails the documents back to the appropriate subcontractor, if needed, to make the required changes. After another round of certifications, the documents are mailed back to the QAR for review. Once the submittal is “approved” or “approved as noted” the general contractor reviews the document one final time and furnishes any needed dimensions or information needed for a purchase order. The general contractor places the necessary orders with the suppliers and/or manufacturers of mails the approved documentation to the subcontractor to place the order. This procedure is illustrated in Fig. 3. Given the overall flow of the three sub-processes contained in the submittal processing tasks, the authors reviewed the details of each tasks to ensure that they were properly classified. The following sections provide a general discussion of this analysis.

Legend

1.10 Package Submittal

Terminator

1.11 Log Submittal

Routing

Processing

Handling Internal Transmission 1.12 SEND

External Transmission

Fig. 1. Submittal preparation process.

One of the first processing task classes are the “product discovery” tasks undertaken by contractors and subcontractors (e.g. Tasks 1.04 and 1.06). Today, the performance of these tasks relies extensively on the past precedent and personal relationships with suppliers and subcontractors. Contractors and subcontractors often assume that the products and materials that were approved on the previous job with a specific client are most likely to be approved in the future. As a result, a frequent starting point for selecting products is the file cabinet of the last project. If those files are not available, an internet search and identification of catalog item will identify materials, products, and equipment submittals. The next type of processing is an “evaluation” task (e.g. Task 1.09). This evaluation task is a critical intellectual activity of the contractor that ensures that components selected not only meet technical performance requirements, but also that the components will be compatible with other selected components. It may be meaningful to

E.W. East, D.R. Love / Automation in Construction 20 (2011) 1070–1078

Owner Representative

Design A/E

Design Consultant

Tenant/ Stakeholder

2.05 Log Receipt

2.12 Log Receipt

1073

1.12 Receive

2.01 Log Receipt

2.02 Routing

2.03 Log Receipt

2.04 Routing

2.13 Review Submittal

2.06 Review Submittal

2.14 Transcribe Review

2.07 Transcribe Review

2.09 Review Submittal

2.15 Log Review

2.08 Log Review

2.10 Transcribe Review 2.16 Review Submittal

2.11 Log Review

2.17 Transcribe Evaluation

Legend Terminator

2.18 Log Evaluation

Processing Routing Handling

2.19 Send

Internal Transmission External Transmission

Fig. 2. Submittal review process.

also note that the identification of deviations may be accomplished anywhere from the manufacturer up to the contractor himself. Since the prime contractor has the final responsibility to identify, resolve, and document deviations, this model assumes that the evaluation of deviation is the job of the prime contractor. Another type of processing activity accomplished by the owner's representative and his team of designers, consultants is to “check” for compliance with requirements (e.g. Tasks 2.03 and 2.09). Reviews by tenants are often required to “select” specific attributes of selected products, such as color or finish (e.g. Task 2.13). On items that require designer and/or tenant review the owner's representative must also “consolidate” the evaluations of several parties into a single communication back to the contractor (e.g. Tasks 2.16). 4.2. Routing tasks Routing, a specialized form of processing tasks, is required to establish the responsible party needed to complete specific tasks. For the contractor, a decision is required to determine if the prime or subcontractor will be providing a specific submittal package (e.g. Task

1.03). The contractor distributes work based on contractual relationships known prior to the requirement to create and submit the required product data. The owner's representative must also decide to whom different types of submittals are routed (e.g. Task 2.02). The designer must also determine who is to review specific “approval” type submittals (e.g. Task 2.07). Once action has been taken the prime contractor determines if the information is to be returned to the subcontractor or kept in-house (e.g.Task 3.02). 4.3. Handling tasks Handling tasks are tasks where the result of processing the “intellectual” tasks are transcribed onto physical documents. These transcriptions occur through the medium of signatures, small forms stamped on documents with hand-written annotations, and general handwritten notes on submittal transfer forms or manufacturer paper documents. Such hand-written documentation has a number of critical drawbacks. Misunderstandings can arise when multiple parties make conflicting handwritten notes. The poor quality of multiple copies of these documents reduces readability of the

1074

E.W. East, D.R. Love / Automation in Construction 20 (2011) 1070–1078

transmissions within an organization. Additional comments on this assumption will be made at the conclusion of this paper.

Sub Contractor

Contractor

5. Implementation and results

2.19 Receive

3.01 Log Receipt

3.02 Routing

3.07 Log Receipt

3.03 Review Evaluation

3.08 Review Evaluation

3.05 Prepare Purchase

3.09 Prepare Purchase

Legend Terminator Processing Routing Handling

3.06 End

3.10 End

Internal Transmission External Transmission

Fig. 3. Initiate procurement process.

documents. Such misunderstandings have resulted in costly change orders [19]. Ambiguities on contractor certified submittals may occur during substitutions of “or equal” products. This is particularly the case when the contractor substitution is not explicitly called out until after installation. The contractor transcribes issues identified when packaging submittals (e.g. Task 1.10). Parties receiving the submittal package for review each transcribe their reviews into a set of ad hoc formats sent to other team members (e.g. Tasks 2.7, 2.10, 2.14, and 2.17). To keep track of suspense dates for every piece of information flowing through the submittal process every party (with the possible exception of some tenant agencies) maintains their own set of submittal logs. For each individual party to a process, logging is typically the first and last tasks of each phase (e.g. Tasks 1.05, 2.01, 3.01). The time required to maintain these log files are identified as handling costs. In the submittal process today, a significant effort is spent copying, logging, and physically moving stacks of paper files. Case study participants indicated that their contracts require between seven and twelve copies of each submittal. Given that projects may have hundreds of submittals ranging from a few pages to hundreds of pages, the volume of paper for a reasonable size project is considerable. Added to the count of initial submittals case study participants also indicated that between 15% and 25% of submittals were rejected which required new packages to be prepared and logged. Handling tasks also identify time required to markup multiple sets of transmittal files forwarded to all required project stakeholders as “preparation” tasks (e.g. 1.10, 2.17, 3.09). Handling errors may result in mistakes in copying marked-up pages and inserting such pages back into follow-up transmissions of such documents. 4.4. Transmitting tasks Today postal mail or other physical courier services move paper submittal documents from one party to another. If the submittal is sent within the United States, it typically takes two to four days for delivery. If the submittal is sent outside the United States, it may take two to three weeks by express courier. One of our case study participants indicated that he personally delivered and received on his trip between his home and the office since it was faster than local postal mailing the submittals. The model identified in this paper assumes that there is no cost to

The authors' prepared the model described in Figs. 1–3 in an MS Excel spreadsheet. Durations for a given path through the network are calculated based on summing the durations of each of the activities in the path paths through the model. Since not all paths through the network reflect typical contract administrative patterns the authors considered two possible sources of authorship of submittals and three possible types of reviews of those submittal packages. The authors considered two possibilities for the initiation of submittals: either the subcontractor or the prime contractor. For each there are three options for owner's action are encountered: (1) acknowledgment of receipt of contractor certified submittals, (2) a designer's review of submittals, or (3) a design consultant's review of shop drawings. This results in a total of six typical paths through the schedule that represent the most commonly prepared submittals on public contracts in the United States. While the process followed by each of these six options is contractually mandated, the technology applied is not. This research considered the use of three different technologies: (1) traditional paper document exchange using postal mail, (2) exchange of electronic documents using electronic mail, and (3) exchange of electronic documents using process-based project extranets. The impact of changing technology is to change the duration of a given task, not to change the activity or the network topology. The results obtained from the model are presented in two sets of data each of the three technology approaches applied to the six different types of contractually required process. The first result provides the total time required for the preparation, review, and initiate phases of the processes. The second result summarizes the total duration of each of the activity types, both value-added and nonvalue added, regardless of the phase in which the task occurs. 5.1. Paper documents and postal mail For the traditional, i.e. paper document, process all processing activities are assigned a one calendar day duration. Routing and Handling activities were assigned a duration of one-half a calendar day. Transmission activities by physical currier services such as postal mail were assigned a duration of five calendar days. The shortest time for the complete submittal cycle was for contractor prepared, certified submittals, with a total duration of 20 days. The longest time to complete the submittal review cycle was found to be subcontractor provided shop drawings that were required to be reviewed by design consultants, with a total duration of 62.5 calendar days. The calculated duration of each of these durations to the six required contractual processes provides the results shown in Table 1. To evaluate this model against typical public construction contract requirements for maximum review times we can see that design consultants were able to complete their reviews in 33.5 days. When the prime Architect/Engineer firm completed a review, it required 21 days. Table 1 Paper document process results. Submittal type

Prepared by

Reviewed by

Expected process duration (calendar days) Prepare

Review

Initiate

Total

Certified Certified Approval Approval Approval Approval

Prime Sub Prime Sub Prime Sub

Certified Certified Architect Architect Consultant Consultant

9.5 21 9.5 21 9.5 21

8 8 21 21 33.5 33.5

2.5 8 2.5 8 2.5 8

20 37 33 50 45.5 62.5

E.W. East, D.R. Love / Automation in Construction 20 (2011) 1070–1078

Contractor certified submittals required 8 days to review. Based on the contractually specified maximum dates for these reviews, the model properly reflects the performance of an effective submittal review team. It is interesting to note that while the focus of the contractual requirements for the submittal process pertain to the maximum time to review, there is a significant time required to prepare the document and initiate procurement. It is interesting to note that, in some cases, the preparation and initiation phases, tasks which are the responsibility of the prime and subcontractors, take longer to complete than the review, which is typically assumed by many practitioners to be the longest duration activity. These durations reflect the most optimistic of situations, one in which all submittals are approved upon first submission. The second result obtained summarizes across activity types. Table 2, shows the results of that summary. It is clear from reviewing these results 2 that non-value added tasks take the majority of the time to process submittals in the traditional method. The total nonvalue added time in the traditional construction submittal process is between 75% and 89% of the total submittal processing time.

5.2. eDocuments and eMail Changes to the traditional model demonstrating the impact of the use of electronic submittals exchanged by email were derived from interviews with contractors rehabilitating a six lane, six-span vertical lift bridge [2]. The bridge, which served as part of a major daily commuter route, served thousands of vehicles per day. Because of the heavy traffic, it was imperative that the project was completed quickly. Before starting, the team agreed upon the allowable file types and the minimum standards for image files were agreed upon to ensure that the quality of the submittals was upheld. The allowable file types included portable document format (pdf), image files (jpg), spreadsheets (xls), and word processing documents (doc). Large files were uploaded to a common file transfer site due to email inbox file restrictions. The contractor credits the resulting accelerated submittal processing for contributing to a four (4) month early construction completion. Team members reported that routing and handling times were not substantially different when using email than when using paperbased submittals. This was due to the need to spend time marking the electronic files, for example to select individual products from a catalog. Since none of the parties shared submittal logs, all parties continued to need to maintain their own databases of submittal dates. Rather than decrease costs, the use of email resulted in an increase in routing cost since the team reported needing to maintain dynamic mailing lists of specific individuals within specific companies. This increase routing time was expected to be within the margin of error of the current one-half of one-day duration for routing tasks. To adapt the models in Figs. 1, 2, and 3 for the transmission of electronic documents by email several changes were required to reflect the process described in Ref. [2]. First, the expected duration for transmission tasks were reduced to one-half calendar day. While the transmission itself may take place more frequently, the receipt of the transmission would not occur immediately since the files would have to

1075

be extracted from email in-boxes. In this case, the construction contractor sent submittals to the designer for processing and not first through the owner. The impact of this process change moved Task 1.12 to the designer's swim lane. In addition, Tasks 2.01 and 2.02 were removed from the model. The resulting durations of each of the three stages of the process are shown in Table 3. The durations for each of the task types are show in Table 4. Reducing the duration of non-value added transmission tasks resulted in a reduction in the time to perform submittal reviews of between 28% and 59% depending on the type of review. Reviews of contractor prepared and certified submittals reduced from twenty (20) days to fourteen (14) days. As would be expected given the large transmission time component, subcontractor shop drawings reviewed by design consultants saw the expect process durations drop from 62.5 days to 25.5 days. 5.3. eDocuments using project extranet The third type of technology evaluated with the value-added model was that of a cloud-computing application, ProjNetsm [16]. The unique aspect of this platform is that it enables the complete electronic exchange of submittal information for the entire project team including from manufacturers' and suppliers to future building tenants. Once submittals are certified, the owner's representative and team including designers, consultants, and tenants may all directly review the documents on-line. Access to a given eSubmittal register is based on assignments, either to individuals or offices, and the specific role of that user or office within the submittal process. Before starting, the team agreed upon the allowable file types and the minimum standards for image files to ensure readability. The concern was to eliminate the creation of PDF files based on scanned copies of faxes, that are commonly found in paper submittals. The allowable file types included portable document format (pdf), image files (jpg), spreadsheets (xls), and word processing documents (doc). The only restriction on file size was the limit associated with end-user internet connections. The only change to the standard submittal processing that the team applied is that one single file was created for each submittal register submittal. Sometimes this resulted in submittal register files that were several hundred pages in length. The process model for the cloud computing case is the same as in Figs. 1, 2, and 3 because the contract requirements for submittals have not changed between any of these projects. There are, however, significant differences in the durations of typical tasks since the shared application supports many of these tasks automatically. For example, routing is accomplished by office and or individual assignments as needed. Once office assignments are provided that office maintains their own mailing lists and does not need transmit changes to all other parties. Routing tasks, could have be set to zero (0) calendar days if the extranet solution had been fully used, however, the team in question routed documents between subcontractor and prime contractor prior to posting against the extranet solution. As a result, the duration of routing tasks was determined to be ½ day per submittal. In the traditional or email case, users are still required coordinate delivered documents against separately maintained submittal registers, all documents are directly attached to their correct submittal Table 3 Email process results.

Table 2 Paper document task type results. Submittal type

Prepared by

Reviewed by

Expected task type duration (calendar days) Process

Route

Handle

Transmit

Total

Certified Certified Approval Approval Approval Approval

Prime Sub Prime Sub Prime Sub

Certified Certified Architect Architect Consultant Consultant

5 5 6 6 7 7

1.5 1.5 2 2 2 2

3.5 5.5 5 7 6.5 8.5

10 25 20 35 30 45

20 37 33 50 45.5 62.5

Submittal type

Prepared by

Reviewed by

Expected process duration (calendar days) Prepare

Review

Initiate

Total

Certified Certified Approval Approval Approval Approval

Prime Sub Prime Sub Prime Sub

Certified Certified Architect Architect Consultant Consultant

5 7.5 5 7.5 5 7.5

7 7 11 11 14.5 14.5

2.5 3.5 2.5 3.5 2.5 3.5

14.5 18 18.5 22 22 25.5

1076

E.W. East, D.R. Love / Automation in Construction 20 (2011) 1070–1078

Table 4 Email process task type results.

Table 6 Extranet process task type results.

Submittal type

Prepared by

Reviewed by

Expected task type duration (calendar days) Process

Route

Handle

Transmit

Total

Certified Certified Approval Approval Approval Approval

Prime Sub Prime Sub Prime Sub

Certified Certified Architect Architect Consultant Consultant

5.0 5.0 6.0 6.0 7.0 7.0

1.0 1.0 1.5 1.5 1.5 1.5

3.0 5.0 4.5 6.5 6.0 8.0

5.5 7.0 6.5 8.0 7.5 9.0

14.5 18.0 18.5 22.0 22.0 25.5

register item. As a result, transmission costs for this case are modeled as having a zero (0) calendar day duration. Given that the information by all parties is kept in a single location and not distributed in a variety of emails the actual processing time of the submittal documents themselves may be reduced from one day to ½ day duration. The only modeling issue with this case was to determine the relative impact of the procedures used on “handling” type tasks. Since the team determined to provide one large file for each submittal package, and since the intrinsic features of the cloud-computing application to collate many small files needed for typical submittal packages, the costs of handling in this model were not changed from either the traditional, paper-based, or email model. The results obtained from the model showing the generalized process durations and work type durations are provided in Tables 5 and 6. As with the email process, a substantial amount of time is saved when comparing the extranet solution to the paper document solution by not requiring the delivery of physical documents. In the email example, a transmission time of 0.5 days was still listed with each transmission task, due to the need of users to match emails and email attachments. Handling costs, such as transcribing document delivery and submission dates are eliminated when using the cloudcomputing approach. Documentation of these “transmission” dates is an intrinsic part of each submittal as the set of submittal documents change status during processing. Since all relevant users have access to the central web server there is no costs to ensure that all relevant users have access to the required submittals. Access is provided through the single routing configuration undertaken by the owner's representative when the submittal register is created. Tracking user actions within the submittal register application is also less cumbersome than in the email example. In email, one has to include “acknowledgment receipts” to document when those later in the process read the message since opening the message is not equivalent to having taken action on the submittal. In addition, tracking of the entire process to include third-party consultants who may have had the documents forwarded to them is not possible. The cloud-computing application allows the entire conversation to be viewed by all authorized project participants. Eliminating transmission and routing costs, and centralizing the location of information upon which decisions are made, decreases the total expected time for processing submittals using extranet technology by between 55% and 77% over the traditional paper-based method of submittal review. Even in a mixed mode of only using the cloud computing application for the submittal review phase, this technology Table 5 Extranet process results. Submittal type

Prepared by

Reviewed by

Expected process duration (calendar days) Prepare

Review

Initiate

Total

Certified Certified Approval Approval Approval Approval

Prime Sub Prime Sub Prime Sub

Certified Certified Architect Architect Consultant Consultant

4.5 5.5 4.5 5.5 4.5 5.5

2 2 4 4 6 6

2.5 3 2.5 3 2.5 3

9 10.5 11 12.5 13 14.5

Submittal type

Prepared by

Reviewed by

Expected task type duration (calendar days) Process

Route

Handle

Transmit

Total

Certified Certified Approval Approval Approval Approval

Prime Sub Prime Sub Prime Sub

Certified Certified Architect Architect Consultant Consultant

4.5 4.0 5.0 4.5 5.5 5.0

1.0 1.0 1.0 1.0 1.0 1.0

3.5 5.5 5.0 7.0 6.5 8.5

0.0 0.0 0.0 0.0 0.0 0.0

9.0 10.5 11.0 12.5 13.0 14.5

resulted in an approximate savings of 40% duration over email processing of submittals. 6. Conclusions The purpose of the value-added analysis is to identify the types of activities required to complete a task. Of the activity types used, only the “processing” type, supports the intellectual work required to select and review, if needed, the proposed submittal. The other three activity types “routing,” “handling,” and “transmitting” reflect the time required to perform work that does not directly contribute to the quality of the finished product. Fig. 4 shows the differences in duration associated with different types of technologies under the exact same contractual requirements. As one might expect, the elimination of the time required to transmit paper by postal mail has a significant impact to decrease the total duration of the submittal process. A feature of this result that is interesting is the difference between the use of email and extranet solutions with respect to transmission time. The key reason for the elimination of transmission time with the extranet solution is that the delivery of submittal documents is determined by centrally managed permissions sets. This eliminates the additional transmission delays associated with the management of multiple mailing lists within each firm. Since the model does not consider submittal rejections that may run as high as 25% of all submittals, the authors consider the results obtained from this model to be very conservative in estimating potential benefits achieved. 7. Recommendations Given the strength of the result comparing traditional process to an e-document process using email or cloud-computing applications, the authors recommend that owners seriously consider enterprisewide adoption of electronic processing of submittals. The analysis can be extended to consider direct costs associated with decreased time of activities on the critical path and with the elimination of reproduction and mailing costs for paper documents. In addition to expected direct benefits to the owner from the use of e-processes, the compiled electronic submittal register provides more accurate facility handover documentation at reduced cost [8]. The authors further hypothesize that virtually all contract administration procedures can be modeled with the task classes illustrated in this paper. Decomposing administrative procedures into tasks classified as “processing,” “routing,” “handling,” “transmitting,” and “terminating” combined with swim-lane diagrams can provide senior leadership and business process analysis quickly developed and easily communicated models demonstrating the business value of technology adoption across a variety of fixed contract administration procedures. As with all survey-based research, a criticism can be made about the applicability of these results outside the context of the small sample used for the research. The advantage of the approach taken, however, mitigates this critique since interested parties can request a copy of the MS Excel spreadsheet from the authors and apply their own

E.W. East, D.R. Love / Automation in Construction 20 (2011) 1070–1078

1077

Fig. 4. Summary results by task type.

deterministic or stochastic distributions to the paths created in the model. This will be particularly useful for large owners who wish to determine the impact of these changes on their entire project portfolios. One example of the limitations of the experimental sample is omission of a case that evaluated the full use of the project extranet. The project team using the extranet that was interviewed did not allow, due to prime contractor's reluctance, subcontractors to directly to upload submittal through the extranet server although the extranet system provided appropriate procedures such direct submission. The expected result when such cases are found can be projected to be the elimination of all routing costs that will further emphasize the wasteeliminating value of project extranets. While the application of technology to streamline existing administrative procedures has been shown to provide significant benefit, this study did not consider the option of changing contractual provisions to support a transition from document-centric submittals to information-centric submittals. It is just such a change envisioned as part of the Specifiers Properties information exchange project [7]. One possible impact of such a change would be to completely automate the selection of contractor certified submittals based building information models provided by specifiers and product data provided directly from manufacturers.

Acknowledgments The U.S. Army, Engineer Research and Development Center, Construction Engineer Research Laboratory in Champaign, IL supported this project under the “Life-Cycle Model for Sustainable and Mission Ready Facilities” project. The authors would like to thank Rama Krishnagiri, Christopher Fink, Mike Canicatti, and Lani Cleghorn for sharing their case study experiences. Finally, the authors would like to thank the anonymous peer reviewers for their constructive comments.

Appendix A. Example interview forms Questions Asked of XXXXX How did the electronic submittal of documents work (program/ method used as well as the order submittals were transferred)? What was submitted electronically versus in writing? What were the benefits of using the electronic submittal process? Did the use of an electronic process allow for an improvement in the quality of the submittals? Typically, how long did it take to receive approval through the written process? Typically, how long did it take to receive approval through the electronic process? Approximately how much time was saved overall using an electronic process as compared to the formal written process? How do you think the electronic submittal process affected the quality of the construction process? How long did it take to transmit a submittal (both electronic and paper)? What was the turnaround time for a submittal both electronic and paper? Regarding file size, were some files too big to be sent electronically? What did you do with these? How did you account for variations in the manufacturer's information? I.E. when things needed to be changed on the documents being submitted? Were there any cost savings from using an electronic process? If so, what? Was a submittal log kept? Was it updated every time a submittal was sent and received? Was this the case for both written and electronic submittals? Were multiple copies of each submittal made? Approximately how much time was spent making these copies? Approximately how many submittals were there?

1078

E.W. East, D.R. Love / Automation in Construction 20 (2011) 1070–1078

Of these how many were submitted electronically? Of these how many were submitted through the written process? Will the electronic process be used in the future? Is it okay to use your company's name in addition to the information gathered? Questions Asked of XXXXXXXX What were the benefits/advantages you achieved from using electronic submittal register? What, if any, are the problems or difficulties encountered when using the electronic submittal register? Did the use of an electronic process allow for an improvement in the quality of the submittals? Typically, how long does it take to receive approval through the electronic process? Is the process faster with the use of the submittal register? By how much? Approximately how much time was saved overall using an electronic process as compared to the formal written process? How do you think the electronic submittal process affected the quality of the construction process? Were there any cost savings from using an electronic process? If so, what? References [1] Robert Amor, Shailesh Jain, Godfried Augengroe, Online Product Libraries: The State of the Art, 2004 http://www.cs.auckland.ac.nz/~trebor/papers/AMOR04A.pdf cited 14-Jan-2010. [2] T.W. Anella, R. Krishnagiri, M. Patel, F. Fuster, Overdrive, Civil Engineering 79 (5) (2009) 60–67 (08857024). [3] J.B. Atkins, G.A. Simpson, According to Hoyle: The Submittal Process, AIArchitect, 2006. [4] Kenneth R. Burkhardt, Trygve Dahl, Mark E. Palmer, Electronic data exchange in the pump industry, Proceedings of the 23rd International Pump Users Symposium, 2007.

[5] Bert deBruin, Anneke Verschut, Erik Wierstra, Systematic analysis of business processes, Knowledge and Process Management 7 (2) (2000) 87–96. [6] E. William East, Julio C. Martinez, Jeffrey G. Kirby, Discrete-event simulation based performance quantification of web-based and traditional bidder inquiry processes, Automation in Construction, 18 (2, Elsevier, UK, 2009, pp. 109–117. [7] E. William East, “Specifiers' Propertites Information Exchange (SPie),” Building SMART Alliance, 2010 http://www.buildingsmartalliance.org/index.php/projects/ activeprojects/32 cited 15-Jan-2010. [8] E. William East, Nicholas Nisbet, “Life-Cycle Information Exchange (LCie), ”BuildingSmart Alliance, 2011 http://www.buildingsmartalliance.org/index.php/projects/ activeprojects/140 cited 5-Jan-2011. [9] Matt H. Evans, “Process Improvement,” On-line Financial Management Course, 2010 http://en.coursgratuits.net/5/lean-thinking.php cited 10-Dec-2010. [10] Michael Hammer, Steven Stanton, How process enterprises really work, Harvard Business Review, Nov 1999 (1999) 108–118. [11] C.W. Ibbs, Product specification practices and problems, Journal of Construction Engineering and Management 111 (2) (1985) 157–171. [12] Ned Kock, Azim Danesh, Paul Komiak, A discussion and test of a communication flow optimization approach for business process redesign, Knowledge and Process Management 15 (1) (2008) 72–85. [13] Dong-Eun Lee, Jonathan Shi, Construction business automation system, Journal of Construction Engineering and Management American Society of Civil Engineers 132 (1) (2006) 88–96. [14] W.R. Mincks, H. Johnston, Construction Jobsite Management, Thomson Delmar Learning, Clifton Park, NY, 2004. [15] National institute of building sciences, Construction Criteria Base (CCB): Unified Facilities Guide Specifications (UFGS), 2009 wbdg.org/ccb/browse_org.php? o=70 (cited 24-July-09). [16] National Institute of Building Sciences, Project Extrant (ProjNetsm), 2010 www.projnet.com (cited 08-Jan-10). [17] National Research Council, Advancing the Competitiveness and Efficiency of the U.S. Construction Industry, National Academies Press, Washington, DC, 2009. [18] R.A. Snowdon, et al., On the architecture and form of flexible process support, Software Process Improvement and Practice 12 (2007) 12–34. [19] P.C. Terry, Communication breakdowns, Practice Periodical on Structural Design and Construction 1 (4) (1996) 108–112. [20] William E. Trischler, Understanding and Applying Value-Added Assessment, Quality Press, Milwaukee, WI, 1996. [21] Lida Xu, Huimin Liu, Song Want, Kanliang Want, Modeling and analysis techniques for cross-organizational workflow systems, Systems Research and Behavioral Science, Systems Research 26 (2009) 367–389.