Data Analytics for Engineering and Construction Project Risk Management

Chapter 1: Introduction to Risk and Uncertainty. This chapter provides: a) general discussion on the types of uncertainties in projects including the examples; we cover theoretical, frequentist, belief-based epistemic, as well as agnostic viewpoints on the uncertainty; we show these viewpoints in context of typical project uncertainties and contrast them against representations of uncertainty in other engineering disciplines; b) summary on the role of knowledge and assumptions in characterizing the uncertainty; we link the discussion on uncertainty to knowledge about the underlying phenomena, the embedded assumptions, and their validity over the course of the project; c) overview on the approaches that relate the risk to the underlying uncertainty; we discuss approaches to the risk-uncertainty relationship in different disciplines, and finally d) discussion on the organizational attitude and viewpoints toward the risk and uncertainty; we cover topics such as value of u

ncertainty (is it always bad?), organizational responsibility towards risk (who should be taking risk, when, and how much?), and the contrast between the decision-theoretic vs. managerial viewpoint on the uncertainty showing the differences that govern the choice of analysis and the methods.

Chapter 2. Project Risk Management Framework. This chapter provides: a) overview of the project systems, their complexity, life-cycle and risk-based decision-making; we define project as a complex system, and its life-cycle in the context of phase-gate process where decisions are evaluated under different objectives and criteria; we emphasize the points where the uncertainty is introduced and when it is reflected in project outcomes; we particularly stress the design and construction/installation i.e. execution phases of a project as this is the key focus of this text; b) outline of the high-level guidelines in conducting risk assessment and management (such as

ISO and PMI approach), the use of "risk language" and common terms in communicating risk (such as SRA glossary of terms), and more detailed description of each step; we particularly emphasize risk identification and assessment as they are the key focus of this text; c) formal definition of risk in projects distinguishing between variability of operations, event driven risk factors, and the combination of the two; also, we discuss risks in context of low probability - high impact and low impact - high probability; we emphasize the role of assumptions and knowledge in formally developing risk statement; and finally d) classifications methods for project risks as they relate to project objectives, their inception and resolution period, relationship to project structure i.e. internal-external, technical-no technical, and other key project parameters. The chapter includes homework examples.

Chapter 3: Project Data. This chapter provides a comprehensive summary on the type and sources of project data, and the methods for data acquisition. The key underpinning of this text is that risk analysis should be driven by data in a mathematically rigorous way; so where can one find such data? This chapter covers project data as they relate to planning and execution phase of the project; more specifically, we discuss data in terms of: a) project phase and system of interest; we contrast available data during planning and estimation vs. data during monitoring and control phase of the project, as well as whether data relates to internal project system (logistics, operations, etc.) or environmental systems (weather, market trends, etc), we define data collection objectives for each of the phase and the system type; b) observed vs. judgement/simulated data, or in other words, whether data is generated by the system and recorded by the participants, or assessed by individuals using their experience, judgements, models, or just gut feeling; we provide a summary of typical sources of observed data such as accounting systems, and other corporate databases on project productivity, cost, schedule, etc., as well as less structured data in terms of archived project documents such as weekly progress reports, audio-visual recordings, project team members communication records, and other types of traceable observations; we also provide an overview of the methods used for eliciting experts' opinion and judgement data: we cover both explicit methods such as workshops, surveys, and Delphi process, including how to aggregate group opinion, and implicit methods such as prediction markets and other simulation games. The chapter includes workout examples and homework problems.

Chapter 4: Probability Theory Background. This chapter is designed to provide enough background so that more advanced theoretical concepts could be introduced. The reader is assumed to have a basic understanding of calculus. However, note

that the objective here is not to provide a formal and mathematically rigor description of the concepts as expected for the students in mathematics and statistics departments, rather it is to provide the introduction using enough rigor that would not alienate the target readers and remain focus on the application. We provide basic introduction to: a) probability types, covering classic, frequentist, belief-based approaches, b) formal definition of independence, conditional probability and Bayes theorem including causal belief networks, c) types and distributions of random variables, d) central moments and its meaning given different types of data - ratio, interval, ranked, categorical; e) central moments of functions of random variables, f) covariance and correlations, g) autocorrelation and time-series, and h) random samples and Monte Carlo methods. The chapter includes workout examples and homework problems.

Chapter 5: Project Planning and Estimating. This chapter is structured in three parts. In the first part we cover the key objectives of risk analysis in context of project planning and estimating (e.g. contingency planning, mitigation plans, and monitoring and control strategies) independent of the stakeholders' perspective and the type of contracting and delivery methods. We identify typically available documents and data as the point of departure, as well as the required outputs for further analysis. In the second part we focus on risk identification and assessment methods distinguishing between three general representations: a) hazard, vulnerability, and consequence; b) probability and consequence, and c) aggregate variability. We further classify those into low-probability high-impact event-driven and high-probability low-impact variability representations. We cover a wide spectrum of identification and assessment models including chain-of-events models such as FMEA, FTA, ET, HAZID, and CCF, and more system-focus models such as AFD/TRIZ, STPA. Further, we discuss the differences between the model outputs and the output requirements for further analysis highlighting the deficiency of some of the most commonly used outputs such as risk matrix and risk registrar. In the third section we focus on the analysis. We provide detailed description of: a) cost analysis which includes linear cost models based on the method of moments, the effect of correlations, mixed variability and event-driven models, network representation, and Monte Carlo implementations; b) schedule analysis which includes Pert approximation, extensions of Pert model to include the effect of correlation, mixed variability and event-driven risks, multiple critical paths, common resources, as well as Monte Carlo implementation, and finally c) joint cost-schedule analysis which includes correlation methods and resource-loaded schedule-cost networks, The chapter includes workout examples and homework problems.

Chapter 6: Project Monitoring and Control. This chapter is structured in two parts. In the first part we discuss the objectives of the risk analysis as it relates to project execution and the process of monitoring and controlling projects (assess the status, detect deviations, and respond). We provide detailed discussion on the available documents and data, including typical indicators as well as their leading vs. lagging nature and the capacity to provide early warning signs. We extensively focus on Earned Value methods and their interpretations in context of cost and time to complete. In the second part we present: a) variability focused methods such as SPC, b) Bayesian learning and update methods, and c) expert judgment methods including prediction markets. Finally, we discuss early warning signs in context on non-probabilistic methods such as knowledge dynamics (assumption validation) and critical slowing down of dynamic systems. The chapter includes workout examples and homework problems.

Chapter 7: Case Studies and Implementation Framework. In this chapter we provide a series of case studies that cover typical situations in which project engineers and managers are faced when dealing with project risk identification and assessment. We distinguish between and cover two types of case studies: the case studies designed to be given as assignments to the students, and the case studies in addition cover implementation framework that is making references to the materials covered in the previous chapters. Listed next are few example situations upon which case studies are developed: a) project manager was given a task to evaluate the requirement for cost/time contingency for a common project for which corporate productivity data is available. How does he/she go about this task? b) project manager needs to work on identifying and assessing technical risks for a complex engineering project, how does he/she go about this given that

data is not available for the considered technology? and c) project engineer is asked to provide update to VP on the status of risk for a major project, how does he/she go about this?