(TCMA-3040) Predictive Analytic Strategy and Capabilities for Project Planning and Controls
Author(s)/Presenter(s): Dr. Manjula Dissanayake, CCP
With the data analytics growing at an explosive rate, there are significant benefits of using this powerful discipline in combination and interwoven into AACE’s TCM framework to create a more sustainable competitive advantage for cost engineering professionals to deliver successful project outcomes.
To harness the power of advance analytics, project owners and contractors need to have an analytics strategy and professionals with related data science skills. This paper will define predictive analytics in the context of project controls and describe an analytics strategy. It will highlight the skills and knowledge areas that are pertinent for project controls professionals to facilitate insight-driven decision making for effective risk management of capital projects.
(TCMA-3046) Construction Project Cost Management – A Data Analytics Based Approach
Author(s)/Presenter(s): Kamesh Tangirala, CCP
Construction industry is a fast paced and continuously evolving enterprise. In the digital age, construction project demands have increased, and so, has the need for new technologies and models that assist in the successful budgeting, planning and execution of these projects. The importance of data in construction is starting to take shape; construction companies, while in the past, relied just on experience to make decisions, now, are looking to data for informed decision-making. This technical paper explores construction project cost management from a data analytics perspective.
The paper explores current trends in data analytics as applicable to the construction industry; analyzes how raw construction cost data can be collected, organized, structured and tailored, to derive patterns, ultimately enabling project teams to make informed and logical cost-related decisions. Further, the paper attempts to lay out a framework to build a working model that can be used to analyze structured construction cost data during the life cycle of a construction project.
(TCMA-3048) Novel Dashboard Using Data Analytics for Project, Contract and Claims Management
Author(s)/Presenter(s): Lucas Bicelli; Diêgo Alves; Henrique A. Takemoto
Nowadays, the project management activities are highly focused on issues regarding project’s budget and schedule tracking. Furthermore, the managing environment demands solutions that can empower the decision-making process and guarantee data reliability. Therefore, implementing resources capable of embed Project Manager’s fast decisions and actions are underlying project management needs.
One of the biggest problems in world and Brazil’s capital projects are claims. The claim risk slightly increases in complex projects, those involving hard access areas, different contractors at the same site, natural conditions, lack of management and productivity deviations. In order to address this need, the companies are investing in technological solutions to thrive their projects, ultimately obtaining better project tracking and preventing capital losses.
This paper aims to propose a dashboard developed with data analytics tool to monitor and control the usage of resources, regarding specifically financial tracking and claim valuation, in order to identify in early stages capital deviations when comparing real vs paid costs, providing red flags to the project management activities, as well as, insights for proper and more accurate decision makings.
(TCMA-3054) Data Analytics to Drive Reporting and Insights for Timely Decisions and Improved Business Performance
Author(s)/Presenter(s): Susan Bomba; Lamis El Didi; Aleshia Ayers
Many large organizations have centralized reporting teams and data analysts that develop a variety of reports for communications with senior management and project and portfolio stakeholders. In an ideal world, the team understands the intricacies of the work and possesses strong data management and analysis skills. The challenges for these teams include having to use data that is often spread in various, disparate systems as opposed to one source, making analysis complicated and time-consuming. Both reporting and data analysis play key roles in influencing and driving decisions and actions that lead to greater value for organizations; however what often results is significant effort to manually curate the data to update recurring reports and less time for analyzing the data to drive key insights for timely decisions and improved business performance. In addition, large organizations cannot pivot as quickly to adopt industry-leading software or data solutions, resulting in a mixture of incompatible systems with different data formats and components. By developing a standardized reporting framework and a robust system architecture with a historical database from a data-driven perspective, an organization can be equipped with an end-to-end reporting structure and process that allows for transparent data analytics, helping project and portfolio stakeholders and leadership reach decisions more quickly and manage their business more effectively.
The paper will focus on four key areas to consider when developing a standardized reporting framework for large organizations: data automation and management, assessing effectiveness and consistency of key performance indicators, layered reporting that caters to the various levels of the target audience, and current systems and tools including the integration of business intelligence and data visualization solutions. It will include the requirements to implement the framework, the development process, and the benefits that it offers.
(TCMA-3070) Redefining Project Information Management in a Post-BIM World
Author(s)/Presenter(s): Marcel Broekmaat
As BIM gains widespread adoption, greater attention must be given to building rigor and richness into the way project information is structured and managed. Traditional construction management tools, information systems and BIM tools are good at creating and tracking basic information about project documents, schedules, materials, labor and detecting clashes and tracking costs. However, data created by these systems is unstructured, divergent and therefore widely underutilized. Creating a common data environment and modernizing project information management is the key to unlocking the true potential value of BIM and project data.
This session will explore the latest innovations in BIM-centric project management including how to leverage new technology, processes and standards for more efficient and explicit sharing of project information and key strategies for eliminating risk associated with interfaces that exist between disciplines. The presentation will share best practices from current and past projects for managing constructability through better project management, including optimizing 3D workflows, creating better and more extensive planning efforts, opening communication and collaboration, and enabling continuous refinement of plan alignment throughout design, construction, and hand-off.
(TCMA-3108) Using Power BI for Project Performance Measurement
Author(s)/Presenter(s): Arun Verma
Power BI is a powerful software that can be used to import project data from multiple sources/databases to analyze information in a graphical interface. Dashboards are created by utilizing Primavera P6 data in Power BI. These reports use real-time data live from the P6 Database and can be accessed from multiple platform such as laptops and mobile devices. Early identification of issues and trends assist project teams to proactively mitigate issues and optimize performance.
In this paper the author will demonstrate practical application of Power BI to provide meaningful project performance metrics. These metrics include Milestone performance, Major discipline Percent Complete, Cash Flow, Earned Value, performance and Schedule Variance analysis.
(TCMA-3119) Data Mining of 3D Model for Construction Planning
Author(s)/Presenter(s): Dong Chen, CCP
Navisworks 3D has been widely used in industrial projects for clash detection, real-time visualization and construction sequence simulation. It becomes one of most important and effective tools for reviewing and communicating information about 3D models between engineering and construction because of its visual effect.
This study attempts to expand the application beyond its visual effect in the area of construction planning. With large-scale geometry data retrieved from 3D models, a number of “fit for purpose” algorithms have been developed. Business intelligence can be obtained in the areas of scaffolding requirement estimate, labor density analysis, and work complexity assessment etc. All of these are key factors to be considered while assembling the project estimate. The current practice is to make judgment calls mainly based on past experience and historical data. This study puts forward a framework of quantifying project site condition attributes with the aim of improving the predictability of project cost and schedule.
(TCMA-3130) Implementing UniModel in an Owner Environment
Author(s)/Presenter(s): Philip D. Larson, CCP CEP PSP FAACE; John B. Newman, CCP CEP
Leveraging the existing capabilities of BIM, and 2D representation, with the addition of metadata (1D + 2D + 3D) we get a UniModel. What is a UniModel? It is a relatively new word that describes an adherence to a process that accounts for the majority (@99.9%) of project costs. Considering that Total Cost Management (TCM) involves determining the quantity of work, including costing and pricing, based on risk/profit; using a combination of a sophisticated library of cost items with detailed resources (Material, Labor, Equipment, etc.), intelligent assemblies, sub-assemblies, plus standardized cost coding structures; will allow cost professionals to more efficiently provide valuable cost data, predictive and historical. The Cost Engineering staff at the Regional Transit Authority (RTA) in the Pacific Northwest, aka Sound Transit, which services Pierce, King and Snohomish Counties in Washington State, have successfully employed part of this process with their Unit Cost Library (UCL) used for the ST3 program presented to the voters and approved for @$54B in 2017. There remain challenges that have yet to be overcome. However, a proof of concept (POC) has been established and the future of managing project cost looks bright.
(TCMA-3152) Forensic Schedule Analytics: A Case Study of Mechanized Mine Shaft Sinking
Author(s)/Presenter(s): Dr. Manjula Dissanayake, CCP; Annie Yu
Mechanical excavation systems have significant advantages over conventional drill and blast methods for shaft sinking, including higher schedule performance, cost savings and enhanced level of labor safety. Some breakthrough technologies have been implemented in recent years in shaft sinking, with vast amount of performance data capture is being automated. With the introduction of technology, project risk profiles are changing.
This paper presents a forensic schedule analytics of a recently completed multi-billion shaft sinking project. It will describe the process of data extract, transform, and analysis; specific tools, techniques, and methods; and how to successfully communicate as planned vs. as built schedule results highlighting the changing risk profile.
(TCMA-3205) Using Big Data Analytics for Project Estimating and Forecast: Part I – Engineering
Author(s)/Presenter(s): Dr. Mohamed E. El-Mehalawi
This paper is introducing a project controls approach that is not dependent on the traditional combo of critical path method and earned value analysis.
This case study is the application of machine learning predication algorithms to project engineering workflow. In this case, more than 100,000 records of data are being used from multiple but similar projects. It is assumed that every document goes through 3 stages like IFA, IFB, and IFC (Issue For Approval, Bidding, and Construction, respectively). Machine learning algorithms are applied on this data to predict the duration between different stages, how long it takes totally, the impact of different engineering disciplines, the location of the office performing the engineering, the number of revision within a single stage, etc.
This machine learning model helps performing organizations to predict the duration and efficiency of engineering and design phase based on discipline and office location. This helps in estimating cost, productivity, contingency, and expected finish dates for engineering. The model provides an estimate for the each engineering drawing and a range for variability. The expected value is the estimate and the variability can be used for calculating the budget contingency and schedule buffers.
(TCMA-3204) Using Big Data Analytics for Project Estimating and Forecast: Part II
Author(s)/Presenter(s): Dr. Mohamed E. El-Mehalawi
This paper is an extension to part I of the same title. It employs machine learning algorithms for estimating future projects and for forecasting current running projects.
This case study is the application of machine learning predication algorithms to project procurement, fabrication, and construction. In this case, more than 100,000 records of data are being used from multiple but similar projects. For every unit of fabrication and construction such as a pipe spool, a cubic yard of concrete, ton of steel, and a mile of cable, there is a set of data collected such as duration of fabrication, duration of installation, number of craftsmen worked on it, duration of testing, etc. Also, for each data element there are attributes like the date, location, weather condition, green-field or brown-field, and remote site or not.
Machine learning algorithms are applied on this data to predict the duration and the man-hours required for each task using advanced statistical analysis. The model provides an estimate for each installed unit and a range for variability. The expected value is the estimate and the variability can be used for calculating budget contingency and schedule buffers.
Author(s)/Presenter(s): Christopher W. Ronak
This paper demonstrates the importance of regularly and accurately calculating forecasted cost-to-complete for major construction projects. The objective is to assist project controls professionals in achieving better cost control and oversight of their projects; and to deliver accurate projections of cash flow to all project stakeholders. Calculating forecast-to-complete is a critical aspect of project controls. It demands a careful process of budgeting, data gathering, progress measurements, change order management, time-phasing and detailed forecasting, to achieve a reliable result. This presentation will dive into the processes and methods required to be able to deliver consistent, accurate results for predicting remaining project costs over a timeline, and early identification of critical issues. What’s equally significant to this method, is that it is practical and achievable; and the result of working with many project controls professionals in real-world applications, in a variety of industries over the past 10 years. This method stands out from previous methods in that it strategically leverages both distributed and automated techniques for capturing and processing key data for feeding the forecasting and project controls engine.
(TCMA-3266) Detecting Relationships in Cost Data Sets
Author(s)/Presenter(s): Dr. Xiuzhan Guo
We are entering the era of big data. Cost data sets might have multiple problems, such as, errors, variations and missing data on the information, differences in cost data captured and maintained by different databases, etc. As the volume and velocity of cost data grows, inference across cost networks and semantic relationships between entities becomes a greater challenge. It is clear that the relationships between cost drivers can provide important insights for developing cost estimating relationships and managing cost risks. In this paper, we shall discuss the importance of detecting relationships among cost data sets and in a cost data set and provide a possible solution of detecting such relationships for cost data and its applications.
(TCMA-3278) Developing and Implementing Visual Dashboards Using P6 Data
Author(s)/Presenter(s): Gino Napuri, EVP
This presentation will focus on how to create enticing and attractive business intelligent dashboards to share with colleagues and clients using data from the most used scheduling tools.
The author, working for an ENR Top 50 Program Management Firm ranked at #14 in June 2018, have experience in large program scheduling, with multiple Contractors, and have developed a workable and efficient method of handling the program scheduling reporting using dynamic and interactive business intelligent analytics tools, like Power BI, Tableau and Google Data Studio. The sharing of multiple effective and successful implemented dashboards will tempt the audience to use this new technology in their everyday work and improve client's satisfaction when it comes to visual reporting of project scheduling.
(TCMA-3280) Integrating Analytics and Storytelling to Present Cost Forecast Changes
Author(s)/Presenter(s): Sunny Goklani
For owners and contractor alike - even with the best cost control practices at hand, it is a difficult task to communicate the monthly cost forecast change drivers, to project executives, through a simple visual story on a piece of paper. This paper aims to solve this problem through applying analytics and offering standardized modules.
With an experience of having driven this practice across a portfolio of construction and remediation projects, the author, through this paper, proposes issuing a monthly Cost FCD (Forecast Change Dashboard) as a medium to communicate bottoms-up analytics to the project executives. The FCD structure will consist of a summary space followed by 5 drag-and-drop modules, and the paper itself will provide standard trackers for ~10 modules to pick from (each tracking PTD performance and derived forecasting, to instill mgmt confidence). Examples: Modules for Labor effort forecasting, risk evolution, commitment analyses, change mgmt, etc.
The Dashboard and module trackers are Excel based, but a Power BI application will also be demonstrated for advanced analytics users. In the age of big data and confusing storylines, the author offers FCD as a way to apply data analytics and story-telling to resolve the age-old problem.
(TCMA-3284) A Case-Retrieval Model for Estimating Design and Engineering Efforts
Author(s)/Presenter(s): Ahmed Abdelaty; Dr. David Jeong
Accurate estimation of engineering and design efforts and their associated costs play a vital role in authorizing funds and controlling budgets during the project delivery process. In many owner organizations and specifically public transportation agencies, a simple rule of thumb such as % of the estimated construction cost has been widely used as a method to allocate the budget for design and engineering efforts, which has led to poor accountability and poor transparencies of funding decisions. To overcome this problem, there have been some studies which tried to develop and use statistics based estimation models using historical data. However, the poor quality of the historical data and the small number of data points have caused poor performance of those models.
This study proposes an alternative to those models using real data obtained from a state transportation agency. A case-retrieval approach is developed to make logical inference for estimating preliminary engineering efforts for a new project. This approach uses a novel similarity scoring methodology to retrieve the most similar projects to make reasonable judgement based on past similar projects. Estimators in public agencies can use the proposed approach to quickly estimate the preliminary engineering efforts in a transparent and reliable way.
This study proposes an alternative to those models using real data obtained from a state transportation agency. A case-retrieval approach is developed to make logical inference for estimating design and engineering efforts for a new project. This approach uses a novel similarity scoring methodology to retrieve the most similar projects to make reasonable judgement based on past similar projects. Estimators in public agencies can use the proposed approach to quickly estimate the preliminary engineering efforts in a transparent and reliable way.
(TCMA-3285) Evaluating Construction Claims Through Effective Information Governance and Advanced Data Analytics
Author(s)/Presenter(s): Trent Williams
Construction claims are often submitted without sufficient back-up information, but extensive documentation is vital to prove the validity of claims. Even for construction projects that thoroughly document issues, it can be difficult to locate, integrate, and analyze information appropriately to evaluate issues and claims. The need for Information Governance, including comprehensive, centralized records management, creates an opportunity to adopt document management platforms and advanced analytics. This paper explores the unique implications of evaluating claims through Information Governance and advanced data analytics. Total cost management professionals, project owners, and other stakeholders can utilize this two-pronged approach to prepare for potential claims, effectively and efficiently review and analyze claims, and, if necessary, prepare for litigation.
By utilizing a centralized document management system, our team searched for and quickly located relevant information amongst a large population of project records, enabling successful evaluation of more than 100 claims totaling over $200 million. Our team evaluated all claims in a short time frame with minimal personnel. Furthermore, our document management approach and claims management dashboard supported our overall strategy to reduce claims processing time, increase resource efficiency, and reduce overall client exposure.