Content uploaded by Denis F. Cioffi
Author content
All content in this area was uploaded by Denis F. Cioffi on Jul 07, 2016
Content may be subject to copyright.
A preview of the PDF is not available
Uncertainty in Task Duration and Cost Estimates: Fusion of Probabilistic Forecasts and Deterministic Scheduling
Abstract
A model for project budgeting and scheduling with uncertainty is developed. The traditional critical-path method (CPM) misleads because there are few, if any, real-life deterministic situations for which CPM is a great match; program evaluation and review technique (PERT) has been seen to have its problems, too (e.g., merge bias, unavailability of data, difficulty of implementation by practitioners). A dual focus on the distributions of the possible errors in the time and cost estimates and on the reliability of the estimates used as planned values suggests an approach for developing reliable schedules and budgets with buffers for time and cost. This method for budgeting and scheduling is executed through either simulation or a simple analytical approximation. The dynamic buffers provide much-needed flexibility, accounting for the errors in cost and duration estimates associated with planning any real project, thus providing a realistic, practical, and dynamic approach to planning and scheduling. DOI: 10.1061/(ASCE)CO.1943-7862.0000616. (C) 2013 American Society of Civil Engineers.
... CPM has been commonly used in practice because of its ability to explain the dependency relationships between activities, finding critical activities or float times, and has been widely used by most project management software. However, for more complex or large projects, CPM had been extensively criticized as inadequate in dealing with uncertainties in construction projects [8,9]. Uncertainty in construction projects is an interesting issue that should be considered in project scheduling. ...
... The program evaluation and review technique (PERT) and Monte Carlo simulation (MCS) are probabilistic methods that have been proposed as a supplement to the CPM to operate with uncertainty in construction projects. Activity durations are viewed as random variables that can be represented by probability distributions in both PERT and MCS [9][10][11]. Furthermore, to produce the probability density functions, probabilistic methods require historical project information. ...
Project scheduling is one of the most essential processes and plays a critical role in determining the success of construction projects. The reliable time contingency enables project planners to effectively address uncertainties and various types of risks that may affect the project duration. The traditional scheduling technique such as deterministic methods and probabilistic methods may not be suitable when used for planning construction projects with uncertainties and risks. This paper proposes a new framework that integrates risk management into project scheduling to establish a more reliable project schedule. The proposed model adopts the adaptive neuro-fuzzy inference system (ANFIS) called the ANFIS-TOOL to model the possibility of risk occurrences integrated with project scheduling in terms of risk lag time (ANFIS-RLT) to generate risk-integrated project duration (RPD). The learning capabilities of adaptive neural networks are utilized to adjust the parameters of the model according to the fuzzy rules, aiming to achieve the most suitable representation of the model. In a real-life case study, a sheet pile wall with a temporary bracing system (SPBS) was applied to demonstrate the application of this technique. The root mean square error (RMSE) was used to validate the accuracy of the model before being applied to real construction projects with a high degree of accuracy. This model can be used as an excellent tool to generate a more reliable schedule baseline in construction projects.
... Researchers have also taken interest in including the cost of activities in the formulation of buffer sizing (Zhang et al., 2014). Examples of other attributes of activities are the fractional distance between each activity start time and the project start time (Ma et al., 2015;Zhao et al., 2010), rework safety time for activity (Ma et al., 2021) and the size and distribution of the error associated with activity duration (Khamooshi & Cioffi, 2013). ...
... The standard deviation of activity duration Resource constraints Activity attributes Tukel et al. (2006) Unified schedule method The size and distribution of the error associated with activity duration Activity attributes Khamooshi and Cioffi (2013) Attribute (1) the elements of the system, (2) the relationship between these elements and (3) the external environment. It then posits that the behaviour of the system can be best described based on not only the elements of the system but also the properties that emerge from the interactions among these elements, as well as the interaction between the system and its environment (Tramonti et al., 2019). ...
The research on project schedule contingency suffers from three major shortcomings: (1) it only partly reflects the complexity arising from the high interdependencies among project activities; (2) it is predicated on the idea of closed systems, and thus does not address the interaction between a project and its environment; and (3) it does not sufficiently inform industry practices. This paper addresses these gaps by conceptually and analytically linking von Bertalanffy’s open systems theory and critical systems practice (CSP) to schedule contingency. It focuses on buffer management as a common approach to schedule contingency, and accordingly develops a method to determine the size of time buffers based on the three foundational dimensions of open systems theory. Further, this research provides guidelines to explain how the simultaneous pursuit of the four stages of CSP can assist project managers and decision-makers in responding to disruptive events.
... Probabilistic modeling allows for a range of possible estimates for activity duration rather than using a singlepoint estimate for a particle activity duration. Therefore, engineers have started to use probabilistic techniques to model project schedules and the cost [13,14]. In summary, Monte Carlo simulation does not require prior information about the numerical data and can act independently. ...
In this article, Monte Carlo Simulation (MCS) is used to evaluate the proactive management of uncertainty in construction contracts. Risk and uncertainty in construction contracts are underdeveloped in Saudi Arabia. Building construction should be grounded in environmental safety, so this area needs to advance quickly. Methods for calculating risk and making decisions in the face of uncertainty were researched. This research paper examined the uses of Monte Carlo simulation to reduce the risks related to project management, the purchase of long-lead materials, and execution, particularly in the construction industry. The outcomes of the simulation modeling clearly demonstrate the selection of the overarching assessment principle and the adoption of the best course of action, calculate the impact of risk on the project's time and cost. MCS can also be used to calculate alternative strategies to lessen uncertainty. Instead of using a single-point estimate, MCS probabilistic modeling allows for a range of potential estimates for activity duration or activity cost. The decision to select the optimistic options with high costs (to maintain the technical system's dependability) but for construction projects is made when there is uncertainty. For environmental impact analyses, the Monte Carlo method is recommended. Future risk management strategies in the construction industry will be greatly influenced by the ability to use the modified methodology to reduce risk.
... PERT is a pseudoprobabilistic project monitoring and control technique because the way in which it estimates the average project duration is deterministic. Namely, PERT simplistically assumes that the average project duration equals exactly the sum of the critical activities' average durations (Khamooshi and Cioffi 2013;Nelson et al. 2016). ...
Classical scheduling techniques are well-known to underestimate the average project duration, yet they remain widely used in practice due to their simplicity. In this paper, the new Critical Duration Index (CDI) is proposed. This index indirectly allows anticipation for the probability of a project ending late, as well as the average project duration extension compared with a deterministic project duration estimate. The accuracy of two simple regression expressions that use the CDI was tested on two representative data sets of 4,100 artificial and 108 empirical (real) projects. Results show that these regression expressions outperformed the only alternative index found in the literature. Besides allowing enhanced forecasting possibilities, calculating the CDI only requires basic scheduling information that is available at the planning stage. It can thus be easily adopted by project managers to improve their project duration estimates over prior deterministic techniques.
Dalam pengelolaan proyek konstruksi, durasi merupakan salah satu komponen yang penting dan mempengaruhi kesuksesan suatu proyek. Akan tetapi akibat unsur ketidakpastian yang terdapat dalam pekerjaan proyek konstruksi mengakibatkan manajer proyek sulit untuk melakukan prediksi durasi akhir proyek konstruksi secara lebih akurat. Pada sisi lain terdapat metode Earned Schedule yang perhitungannya menggunakan indikator berbasis waktu, yakni Schedule Performance Index (SPI), sehingga mendukung dalam perhitungan prediksi durasi akhir proyek konstruksi. Oleh karena itu, pada penelitian ini akan dirancang sebuah model perhitungan rentang durasi probabilistik akhir proyek konstruksi dengan menggunakan metode Earned Schedule. Penelitian ini menggunakan data berupa kurva S dari 45 proyek yang sedang berjalan di Jakarta dan sekitarnya. Penelitian ini dilakukan dengan menghitung angka rata-rata prediksi durasi akhir proyek (EAC) pada masing-masing data proyek, kemudian dilakukan perhitungan interval batas atas dan batas bawah sebagai rentang durasi probabilsitik akhir proyek. Hasil dari penelitian ini menunjukkan bahwa model perhitungan rentang durasi probabilistik akhir proyek dapat digunakan untuk memprediksi durasi akhir proyek yang bersifat probabilistik.
In a highly competitive market environment, organizations must improve their productivity, reduce production costs and improve management methods, in order to maintain a favorable competitive position. Project factors, such as time, cost and risk are closely linked. If one of those factors is out of control, the progress of the whole project will inevitably be delayed. When the progress of the project is delayed, duration compression measures such as crashing or fast tracking are usually employed to ensure the backward schedule returns to the original plan. However, the implementation of duration compression measures increases the cost and risk of the project. In most previous study of the critical path, only activity time is involved, and cost and risk are not considered. In the present study, a mathematical model is built to solve the optimal duration compression scheme objectives of the minimization of risk and cost. The model is verified by two cases, and the best solution is obtained by using LINGO software and Excel solver. The mathematical model is found to provide the best duration compression scheme for the project, with the least increases in cost and risk.
Earned value management (EVM) is a classical project monitoring technique that is widely used in construction projects. Due to its simplicity, this technique suffers from limitations due to its discrete nature – activity durations, costs, and progress are gathered only at update points with no information in between. These limitations preclude EVM from being easily implemented on some project types (e.g. repetitive projects) and in conjunction with some planning techniques (e.g. linear scheduling), where information continuity is both possible and desired. Therefore, in EVM is reformulated based on singularity functions (SF). SF are a type of expressions that can be easily concatenated to model continuous inputs at the activity-level. SF are also additive so as to immediately yield project-level performance information. It is demonstrated how the complete theory of EVM is newly expressed in SF. This offers several advantages: (1) EVM metric axes can be easily swapped (allowing exact calculation of modern metrics such as Earned Schedule or the p-factor); (2) activity progress data can be inserted at any frequency as the available data allow; and (3) short-term project duration and cost forecasts are directly possible for the first time. These advantages are exemplified on a real construction project. Finally, it is discussed how the new formulation with SF produces more accurate project duration and cost estimates compared to the former discrete EVM on real construction projects.
The COVID pandemic has given rise to the necessity of social distancing regulations, which has brought the importance of workspace management on the construction site to an unprecedented level. Understanding and visualizing the interaction and tradeoff among space, time, and workforce is critical for construction managers to schedule and deliver projects on time. Therefore, the objectives of this research are to investigate how the critical path method (CPM) and Takt-time planning methods utilize space, time, and workforce differently, develop a tool to visualize the space-time-workforce interactions, and investigate the space-time-workforce tradeoff based on different managers' preferences. This research selected a high-rise office building project and collected 889 sets of productivity data of five specialty trades. The research built a simulation model to investigate productivity and project performance under 267 scenarios of different combinations of the three resources. A dynamic tool was then developed to visualize workspace, time, and workforce interactions. Finally, a Choquet integral-based evaluation and decision tool was developed. The simulation results show that the Takt-time planning method can reduce up to 80% of workspace overlap compared with the actual production plan with less than 20% of duration extension. The contributions to the body of knowledge are (1) creating a visual framework for managers to understand the interaction and tradeoff among space, time, and workforce quickly and accurately, and (2) developing an innovative Choquet integral approach for managers to evaluate planning strategies according to project preferences. The framework and analysis method can be adapted to other construction projects to assist managers to visualize and optimize the space-time-workforce tradeoff under uncertain project drivers.
This article presents results from the first statistically significant study of cost escalation in transportation infrastructure projects. Based on a sample of 258 transportation infrastructure projects worth $90 billion (U.S.), it is found with overwhelming statistical significance that the cost estimates used to decide whether important infrastructure should be built are highly and systematically misleading. The result is continuous cost escalation of billions of dollars. The sample used in the study is the largest of its kind, allowing for the first time statistically valid conclusions regarding questions of cost underestimation and escalation for different project types, different geographical regions, and different historical periods. Four kinds of explanation of cost underestimation are examined: technical, economic, psychological, and political. Underestimation cannot be explained by error and is best explained by strategic misrepresentation, i.e., lying. The policy implications are clear: In debates and decision making on whether important transportation infrastructure should be built, those legislators, administrators, investors, media representatives, and members of the public who value honest numbers should not trust the cost estimates and cost-benefit analyses produced by project promoters and their analysts.
Resource-constrained project scheduling involves the scheduling of project activities subject to precedence and resource constraints in order to meet the objective(s) in the best possible way. The area covers a wide variety of problem types. The objective of this paper is to provide a survey of what we believe are the important recent developments in the area. Our main focus will be on the recent progress made in and the encouraging computational experience gained with the use of optimal solution procedures for the basic resource-constrained project scheduling problem (RCPSP) and important extensions. We illustrate how the branching rules, dominance and bounding arguments of a new depth- first branch-and-bound procedure can be extended to a rich variety of related problems: the generalized resource-constrained project scheduling problem, the resource-constrained project scheduling problem with generalized precedence relations, the preemptive resource-constrained project scheduling problem, the resource availability cost problem, and the resource-constrained project scheduling problem with various time/resource(cost) trade-offs and discounted cash flows.
Project schedule development is an iterative process that determines the planned start and finish dates for project activities. Within this process the anticipated project duration is also determined. The primary aim of this paper is to stress the importance of historical information as well as the right distribution selection in activity duration estimating, by comparing the different outcomes of scenarios when historical information is or is not used. The scenarios were implemented on 20 small- and medium-sized construction projects. Both the Programme Evaluation and Review Technique (PERT) and Monte Carlo Simulation (MCS) are used and the results produced under four different scenarios are compared. The first scenario uses standard PERT, while the three remaining scenarios are simulation approaches differentiated in the use of the right distribution or not and in the use of historical information or not. The findings validate that MCS is superior to PERT and moreover expose the difference in the results when the suitable distributions are selected based on accurate historical information compared to when historical information is not available.
The development and use of risk analysis techniques in relation to oil field development is described. CATRAP, a cost and time risk analysis program, is presented, and two examples of its use with oil-related projects are described. A further example is given of a project for which CATRAP is not appropriate, and for which a specific probability model is required.
Recently, a Norwegian government report on the cost overruns North Sea projects was presented (NOU 1999:11). It concluded that there was a 25% increase in development costs from project sanction (POD, Plan for Operation and Development) to last CCE (Capital Cost Estimate) for the 11 oil field projects investigated. Many reasons like unclear project assumptions in early phase, optimistic interpolation of previous project assumptions, optimistic estimates, and underestimation of uncertainty were given as reasons for overruns. In this paper we highlight the possibility that the cost overruns are not necessarily all due to the reasons given, but also to an error in the estimation and reporting of the capital expenditure cost (CAPEX). Usually the CAPEX is given by a single cost figure, with some indication of its probability distribution. The oil companies report, and are required to do so by government authorities, the estimated 50/50 (median) cost estimate instead of the estimated expected value cost estimate. We demonstrate how the practice of using a 50/50 (median) CAPEX estimate for the 11 projects, when the cost uncertainty distributions are asymmetric, may explain at least part of the “overruns.” Hence, we advocate changing the practice of using 50/50 cost estimates instead of expected value cost estimates for project management and decision purposes.
The view is expressed that alternative concepts of good project management should be sought that are not necessarily based on traditional Western ideas. Although Western thinking has had a strong influence in the past, it has been shown to be fallible, and the experiences of other countries should be considered. One alternative, formulated in Scandinavia, is described.
Everybody knows Parkinson's Law. Nearly everybody quotes favorite cases to warn, amuse, or arouse. The Parkinson phenomenon is recognized by managers and newsmen, by bureaucrats and their critics. The readers, even the youngest ones, can't help knowing about PERT. Some taught it, others studied it; some practiced the art, others were practiced upon. Parkinson's Law and PERT were launched internationally, and indepently, in 1957. Few afficionados of either know that Parkinson and PERT address the same subject. This close kinship deserves a popular exposition. As the two recede into history and merge into the common stream of managerial lore the perspective gained makes this task urgent.
This paper presents a mathematical model for calculating the budgetary impact of increasing the required confidence level in a probabilistic risk assessment for a portfolio of projects. The model provides a rational approach for establishing a probabilistic budget for an individual project in such a way that the budget for a portfolio of projects will meet a required confidence level. The use of probabilistic risk assessment in major infrastructure projects is increasing to cope with major cost overruns and schedule delays. The outcome of the probabilistic risk assessment is often a distribution for project costs. The question is at what level of confidence (i.e., the probability that budget would be sufficient given the cost distribution) should be used for establishing the budget. The proposed method looks at a portfolio of such projects being funded by the same owner. The owner can establish a target probability with respect to its annual budget. The model can help the owner establish confidence level for individual projects and also examine the effect of changing the confidence level of the portfolio budget on the budget and the confidence level of individual projects. The paper is relevant to practitioners because it provides a methodology for establishing confidence levels by the owner agencies in the emerging field of cost risk assessment for infrastructure projects. A numerical example is provided using actual transit project data to demonstrate the model application.