Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
31 chapter five CONCLUSIONS This study examined construction staffing practices at state transportation agencies (STAs) and selected non-STA trans- portation organizations using an online survey tool, site visits, and a review of STA literature on construction staffing. This project developed several findings and recommendations for future work in response to the work summarized here. GENERAL FINDINGS The data analyzed in this project offer a number of signifi- cant general findings related to STA construction staffing for highway construction projects: ⢠Few STAs reported having formal systems to estimate construction staffing needs for highway construction projects. Of the 40 STAs contacted regarding formal con- struction staff forecasting methodologies, seven states indicated that they use some type of formal system to estimate construction staffing needs for future projects. ⢠STAs are managing larger roadway systems with fewer in-house staff than they were 10 years ago. For the 40 STAs that responded to the survey, between 2000 and 2010 state-managed lane-miles increased by an average of 4.10%, whereas the number of full-time equivalents (FTEs) decreased by 9.68%. When FTEs are normalized across the managed road system, the responding trans- portation agencyâs FTEs per $ millions of disbursement on capital outlay decreased by an average of 37.26%. ⢠The types of construction staff forecasting methods employed by STAs are diverse and widespread in their methodology. The forecasting methods range from sim- ple staffing heuristics based on generic project types to multi-variate regression models developed from his- torical project data. These methods also varied in the processes used to estimate staffing numbers with some using work type and others using total project cost to estimate staff requirements. ⢠The two most cited factors by responding STAs for increas- ing construction staffing requirements for a project were poor quality plans, specifications, and cost estimates, and an accelerated construction schedule. Other factors that increased staffing requirements for construction administration and construction engineering personnel differed from those for construction inspection. Con- struction engineering and construction administration staffing requirements were supplemented by increased third-party coordination efforts. Construction inspection personnel requirements were increased by expanded environmental mitigation requirements. ⢠Few factors were identified that tended to decrease con- struction staffing requirements for highway construc- tion projects. The lone exception was that increased experience for construction inspectors and contractors reduced the number of construction inspection person- nel required. However, it is important to note that the survey did not specifically collect data on factors that could decrease construction staffing requirements. ⢠Outsourcing of construction personnel is more common now than reported in previous studies. Ninety-six percent of survey respondents noted using consultant personnel to meet staffing needs in construction administration, engineering, and inspection. The most common reason cited for the use of consultant labor was inadequate in- house construction staff. ⢠The adoption of mobile information technology (IT) within STA construction organizations appears to be limited. Less than 30% of responding agencies reported smart phone use and less than 15% reported using tablet computers among their field personnel. Of those using mobile IT, 60% reported no increase in user productivity from the mobile devices. Limited data were available to identify why adoption has been slow; however, the data collected indicated a lack of system support from STA central IT departments and the limited availability of mobile applications specific to highway construction. COMMON CHARACTERISTICS OF CONSTRUCTION STAFF FORECASTING SYSTEMS For STAs that are interested in developing a construction staff forecasting network for their own agency, the systems examined in the current work share a number of common char- acteristics to be considered when developing a new system. ⢠A timeline for the construction staffing forecast. Although the timelines for the systems differ, all the systems exam- ined in the current work base their staffing estimates on a specified analysis period. The systems in use at the North Carolina Department of Transportation (DOT) and North Dakota DOT forecast staffing needs for a single proj- ect. The system in use at the Michigan DOT focuses on projections for a single calendar year. Those systems in use by the Utah DOT, Texas DOT, and California DOT (Caltrans) forecast at a more strategic level over several
32 decisions, budgeting decisions, and project selection. Such a tool could be used to estimate the number of FTEs needed to execute a project portfolio and management could then adjust resources accordingly (hire new employees, bring in consul- tant labor, employ interns, etc.). However, the tool could be used as a decision aid in more areas than simply adjusting human resources. If a forecast shows that a program plan results in unsustainable variations in FTEs, the individual project schedules could be adjusted to distribute the projects more evenly and allow for a more sustainable staffing plan. In addition, if a spike in human resources is identified, STA management could review the assigned duties and responsi- bilities of construction personnel and temporarily (or perma- nently) adjust duties and responsibilities to allow the existing workforce to cover a larger project volume. The forecasting methodologies used in the current work implement two general approaches: (1) construction staff needs based on staffing metrics or (2) construction staff pro- jections based on historic project staffing data. It is not pos- sible to identify which method is superior; therefore, future research could assist in the development of forecasting tools for construction staff. Given the variation in the resources, organizational structure, project delivery methods, project volume, project type, and human resources across STAs it is unlikely that a single forecasting tool could be applied to all STAs. However, future research could address some common issues related to forecasting construction staff for highway construction, including: ⢠What is the most accurate method to forecast construc- tion staffing for future projects? Recent work in examin- ing the accuracy of contract time determination at STAs demonstrated the lack of accuracy in at least some these tools for estimating contract time. Part of the reason for this inaccuracy was the lack of validation of the tools. Of the construction staffing forecasting tools examined in the current work, none reported any validation efforts to date. For the tools that use staffing metrics, do these metrics reflect the adequate staff requirements for a given project? Do they overestimate or underestimate the staff needed? For the tools that rely on regression analysis has the accuracy of these systems been tested against actual projects? ⢠What factors have the greatest impact on the staffing requirements for highway construction projects? Any forecasting tool in current use is based in some way on an average or typical project; however, there are many variations in highway construction projects. Identify- ing the factors that cause the largest fluctuations in construction staffing requirements would allow staff- ing for a project to be adjusted based on the unique project characteristics. This work should also consider how project performance metrics (e.g., achievement of schedule milestones, accelerated construction require- ments, contractor evaluations) and construction per- sonnel experience factor into the construction staffing years. It can also be noted that of the systems examined only the Utah DOT and Caltrans systems began formally tracking construction projects during the design phase. It can be noted from the systems examined that the longer the analysis timeline, the more complex the system, and likely the more resources required for development and maintenance. ⢠Some form of project schedule is needed to estimate staff- ing needs. None of the examined forecasting systems reported developing a critical path schedule as part of their methodology. However, each system used some formal or informal method to estimate project dura- tion, with some systems including some generic type of activity. The North Dakota DOT system estimated proj- ect duration based on activity type and construction- miles of the work zone presumably developed from the analysis of historical data. The Michigan and Texas DOT systems included an estimated start and end date for the projects. The Utah DOT system uses project mile- stones throughout the design and construction phases and Caltrans uses assumptions about the durations of basic activities. ⢠Some type of connection between staff requirements and the work performed is needed. The Kentucky Transpor- tation Cabinet, West Virginia DOT, and North Dakota DOT publish recommended staffing standards for dif- ferent types of work. The Texas DOT uses a regression analysis of historical data to estimate staff requirements for a given volume of work as well as some assump- tions as to the staff requirements based on certain project characteristics. Caltrans uses assumptions about the vol- ume and type of work that can be managed by a single individual. The Michigan DOT system does not explic- itly describe how staff requirements are linked to work type; however, its user must make this connection when entering man-hours into the system. Although most of these systems use some type of historical data or published standard for staffing levels, these data are not to be used without taking into account the cur- rent project or project portfolio. Relying on historical data can lead to a self-fulfilling prophesy, where since a project provided a certain level of staff that staff is used regardless of whether more or fewer people are needed on the project. FUTURE WORK This current work highlights the lack of widespread use of formal construction staffing methodologies across STAs. This does not imply that STAs are not performing some type of construction staffing analysis at either an informal level or as a discrete, periodic planning exercise. However, as these agencies continue to be tasked to manage larger infrastruc- ture systems with fewer employees the need for an accurate estimate of construction staffing personnel will be critical. A tool that accurately forecasts construction staff over both the short and long term would improve personnel management
33 staffing requirements could identify design areas that might receive increased attention during the design pro- cess to minimize staffing requirements during the con- struction phase. ⢠How can currently available mobile IT be used to improve FTE productivity at state transportation agen- cies? Mobile IT devices such as smart phones and tablet computers are being used by private business to improve staff productivity. However, the results of this synthesis show that improvements offered by mobile IT devices to STA personnel have been minimal. As previously noted there are several potential hindrances to the impact of new IT on staff productivity. There may also be lim- ited availability of IT infrastructure at the project site. Discussions with several STAs indicated reluctance by their central office IT groups to allow the use of wire- less technology owing to concerns with data security. The limited availability of mobile applications specific to the transportation construction industry may also contribute to the low acceptance of IT. Finally, there are different levels of compatibility between current tablet operating systems and existing STA IT systems that may prohibit the use of certain tablets. Additional work in this area could identify barriers to the useful- ness of mobile IT platforms in STA construction divi- sions and develop recommendations for how mobile IT can be used to improve STA construction personnel productivity. Addressing these issues could aid in the development of a construction staff forecasting system that is accurate, easy to use, and straightforward to maintain. A draft NCHRP research needs statement to address these issues is included in Appendix D. requirements for a specific project. The survey results indicated that a more experienced contractor lowers STA construction staffing requirements; therefore, it would appear that the average experience of the con- tracting fleet for specific STA regions would need to be included in the staff forecasting system. ⢠What is the ideal construction staffing level for a given type of project? With reductions in the number of STA in-house personnel across the country (and additional reductions possible in the near future) a balance must be struck between the duties and responsibilities of con- struction staff, the volume of work being managed by a single person, and the budget constraints of the STA. Too many staff assigned to a single project wastes resources and can lead to inefficiencies. Too few staff could lead to cost overruns, project delays, and final construction products that do not meet desired quality standards. Part of this investigation would need to examine methods to increase the productivity of existing staff whether through adoption of technology or modifications to the current job responsibilities of each position. ⢠How can plans, specifications, and project administration processes be modified to improve construction staff effi- ciency? The survey results identified poor quality plans, specifications, and estimates as the most commonly cited contributors to increased construction staffing require- ments for highway construction. Future research in this area could identify specific plans, specifications, and project administration deficiencies that lead to the great- est increase in project staffing requirements. For example, recent work on analyzing the causes of change orders at the Kentucky Transportation Cabinet identified guard- rails as a high change order risk work item. Research into design elements that lead to increased construction
