What are the key factors and assumptions that affect the accuracy of parametric estimation?

Use of good quality data is critical for accurate parametric estimating in civil engineering, particularly during the early stages of design maturity. Investing in high-quality data collection and analysis can help to ensure that projects are estimated accurately. If the parametric estimate for a construction project is based on outdated cost data / inaccurate material prices, the resulting estimate may not reflect the current market conditions, and the cost may end up being under budget. The quality of data used in parametric estimates can also impact the level of uncertainty associated with the estimate. Using good quality data help mitigate uncertainty and improve the accuracy , providing greater confidence in the estimated cost.

To provide reliable outcomes parametric models need to consider more than just the outcome data from previous projects. Models that also understand additional factors around the organisational capabilities, economic conditions and associated systems will also provide more robust outcomes. With the introduction of new materials, processes and construction techniques measures of technology readiness also need to be considered. Good examples of mature parametric models include TruePlanning from PRICE Systems and The SEER Suite from Galorath.

In the defence and security domain the requirement in the data is quality and quantity. The issue is due to these types of projects are long in duration, many years until you get the information, and cutting edge technologies to overcome the threat, there are only a few different aircraft carriers! The data you need to generate a parametric model is technical, schedule and cost. All the data needs to be outcome data, not forecast data from bids. The technical data needs to be normalised to a single type; Kg, m, N. When you generate a parametric database consider trading cost data with organisations to grow the quantity and alternative products. “I will give you this historical data, what can you offer me?”

Parametric estimation is a statistical technique used to predict project costs/durations. The process relies on model selection, which involves analysing historical data to identify dependent and independent variables. Model comparison/testing is crucial, with the model with the highest R-squared value being the most suitable. Parameter adjustment and refinement are necessary to optimize the model's fit to the data. Validation and verification are essential to ensure the model's accuracy and generalisability to new projects. Incorrect model selection leads to significant errors, impacting project planning, resources, and decision-making. Model comparison, parameter refinement, and validation are essential for successful project execution.

One of the major threats to the accuracy of parametric estimates is the presence of risk and uncertainty. Both risk and uncertainty can lead to inaccurate parametric estimates, which can have serious consequences for the project. To mitigate these risks, it is important to conduct a thorough risk and uncertainty appraisal, identify potential risks and uncertainties, and develop appropriate mitigation strategies. This can help to ensure that the parametric estimates are as accurate and reliable as possible, which can in turn help to ensure the success of the project.

Stakeholder involvement Stakeholders are key to parametric estimation success. Their input can improve accuracy but also introduce biases and pressures. To manage stakeholder involvement: Identify and engage key stakeholders early and frequently. Share estimation methods, assumptions, and results to solicit feedback. Negotiate any issues or discrepancies to ensure expectations are met. Engage stakeholders early and often to get their input and feedback on parametric estimates. Manage their expectations to avoid biases and pressures.

Estimation maturity is the degree to which an organisation's estimation process is established/optimised. A higher level of estimation maturity can lead to more accurate, consistent, and efficient estimates and fewer pitfalls. To improve estimation maturity, it should: Define and document their estimation process and practices. Train and educate team members and stakeholders on the estimation process and procedures. Implement and use appropriate tools and techniques to support the estimation process and procedures. Review and evaluate estimation performance and outcomes to identify areas for improvement. Improve estimation maturity by defining and documenting your strategy, using appropriate tools and reviewing/evaluating your performance.

Find a copy of the book "Cost Engineering Health Check: How Good are Those Numbers?". This is a cost estimate maturity assessment based upon the Knowledge Base Estimating (KBE) elements; Data, Tools, People and Process. Within the elements of KBE there are questions of your organizations reality and a reflection of the most mature solutions. So, read the questions and consider the situation, if its not mature then identify justified actions to determine the best outcome. These actions will increase your maturity and competitiveness. High quality estimating leads to accurate budgets, a reduced potential for cost growth, accurate evaluation of risk exposure, and the opportunity to implement effective earned value management (EVM).

Another factor to consider when doing parametric estimates, is the condition of the site, soils, and utility locations. I know when doing parametric estimates, a lot of the "detail" is not readily available, but the issue needs to be brought up to all parties. A site geotech report done early in the planning stages is essential. I have worked on several projects, that have gone into design, then while in design, a soils analysis is done, only to find that the building will need some sort of drilled pier, or as I like to call it "enhanced" foundation system. That can add 20% or more to the project cost.

Obtain a copy of the English (or Chinese) book "Systems Cost Engineering" as this considers the applications of parametric models. Parametric models are flexible tools which bring engineering, scientific and mathematical rigour to cost and schedule estimating, but great tools alone will not keep programs affordable. Tools must be applied as part of a credible process if estimates are to be accepted. Complex projects all suffer from two problems: the project sponsors often struggle to specify the project effectively, and project managers find themselves wrestling with unpredicted cost or schedule overruns. Everyone wants to be successful with the tools they use, so this book is a comprehensive collection of methods with proven success.

Another example is: I worked on a restroom project at a local university, where they were building 5 single use compliant restrooms buildings throughout the campus. In one case, the utility connections were nearby and the costs were reasonable, in another case, the utilities were 500' away through hilly terrain. Even though the buildings were the same, each restroom project varied significantly from one another. Getting this pertinent information early during budgeting, or parametric estimating is critical for project success.

Key factors affecting parametric estimation accuracy include project scope clarity, data quality, expertise in modeling techniques, and environmental conditions. Assumptions on resource availability, technology efficiency, and regulatory compliance also play crucial roles. For instance, in offshore wind projects, accurate estimation hinges on wind patterns, seabed conditions, and equipment performance assumptions. Grid code compliance infeasibility studies rely on assumptions about grid stability and regulatory requirements. Similarly, manufacturing and battery energy storage estimations depend on material costs, technological advancements, and regulatory incentives.