Analytics

Descriptive

In the Descriptive phase, the focus is on "What happened?" Often a deep dive into historic data is the first step, where standard statistical analysis is performed. A key objective is to identify patterns that expose opportunities for improvement and patterns of success that can be transferred to other parts of the organization.

Predictive

Building on what has been learned from past performance and events, organizations enter the Predictive phase. The relationships between data inputs and outcomes form the basis for models forecasting future states or performance, "What could happen?" The typical tool employed during this phase is simulation. A digital twin of a current process is developed and validated with existing data. During model development, opportunities for improvement become obvious and project momentum grows as insights are generated.

Prescriptive

Game changing results occur when an organization passes into the Prescriptive stage, "What should we do?" This is a phase where opportunities for innovation and insight into value-creating activities are discovered. Historically, innovation required substantial upfront investments into design, capital and organizational change with the hope that predicted outcomes would occur. Today, we can weave existing data and subject matter expertise together to create inputs for models exploring new business processes. At this point, any one of, or a combination of many Analytics tools are applied to generate these insights. The potential business growth opportunities and risks can be identified quickly.

Cost or Schedule Range Analysis (CRA or SRA)

For all projects, no matter the size, location, or industry, there are unknowns.  Even though risk and variability are both a subset of uncertainty, we categorize the risk on projects as variability or uncertainty.  Variability refers to a random nature of a process – where the outcomes are random even though the process and its parameters are well understood. Examples might be the variability of labour rates, estimated quantities, estimate material and equipment costs, and productivity.

Uncertainty, in this case, refers to the lack of knowledge about the value of a parameter, or the factors that determine its behaviour. Uncertainty usually applies to factors on projects that are not required to execute the project, but if they occur could adversely affect the cost or schedule or sometimes both. Contingency is an amount of money or schedule added to a budget to account for all the things that haven’t been defined or quantified yet. Uncertainty in a cost estimate is a concept that is readily accepted in the industry no matter what stage a project is defined.  The question, however, is how much uncertainty exists? 

Traditionally and prior to analytical methods to determine contingency, contingency is set as a factor of the total estimate based on experience and the level of project definition. Engcomp still uses this concept to establish a value, but we back it up with a statistical analysis of the estimate that generates a confidence level for the total estimate.  Our Contingency Analysis is a quantitative Monte Carlo approach that seeks to statistically determine the possible outcomes of a project total cost or schedule. 

The results of this Monte Carlo simulation help to define a range of possible cost and schedule outcomes to complete the project, so that the owners can select a budget with a specified level of confidence, that the project will be complete on or under that budget. 

In general, the techniques in our Contingency Analysis encompass four steps:  Model Development, Uncertainty Identification, Analysis, Decision Making. Along the way, we review estimates with our clients, calibrate the planning team and develop a useful deliverable that brings clarity to hazy situations for our clients.

Integrated Cost & Schedule Range Analysis (CSRA)

Integrated Cost & Schedule Range Analysis takes an analysis that focuses on solely cost or schedule to the next level, quantifying the effects of all reasonable risks and uncertainty on a project.  This process takes into account the variability of costs, schedule durations, outside project discrete risk events and their collective impact on the project.  The results of this Monte Carlo simulation help to define cost and schedule contingency required to complete the project for a given confidence level as well as the risk reserve budget required. 

One might question – why do you need to quantify contingency separate from risk reserve on projects?   The answer lies in the control authority of these two pots of money.  Recall that contingency is an all encompassing amount that allows for all the items you couldn’t define or quantify at the given stage of the project. The project manager should be given full control of the contingency to manage and allocate those funds as necessary amongst the project cost centres or work packages.  Risk Reserve is an amount of money that the project carries to cover reasonable project risks that may affect the project cost.  The key here is that this portion of the budget does not account for any items that are required to be built. Risk reserve, usually requires authorization at a more senior level based on appropriate justification for its use. 

The objective of Engcomp’s CSRA process is to provide a framework and methodology to derive a reliable project cost estimates as well as to identify and manage risks associated with the project.

Simulation

Simulation and Optimization are tools that start us down the Prescriptive phase.

One of our favourite Analytics quotes is from George Box "All models are wrong, but some are useful." The future is uncertain, so when we can, models should include this variability. Simulation is the tool that we use to understand and quantify this variability.

Most organizations will create a business plan for a new venture or project. To paraphrase Eisenhower, once the fighting has begun, the battle plan is worthless. In other words, the business plan is based on a static combination of assumptions for a large number of inputs. The likelihood of all assumptions being predicted correctly are small. So, while the intention is good, the business plan is at best a guide at managing the enterprise. To better understand the problem, a technique like Monte Carlo analysis (static simulation) introduces uncertainty on the inputs and assumptions so there is better understanding of what could happen if everything goes great, but also what could happen if nothing does. The real value of Monte Carlo analysis is that it quantifies the likelihood of the various outcomes so you can see the range and likelihood of what future outcomes await.

Simulation can also be used as a design or operational tool. Digital twins of a production environment such as a mine, manufacturing facility, warehouse, or retail operation can all be produced by a dynamic simulation.  These digital twins allow organizations to make "what-if" analyses; exploring how potential operational changes will impact their business without any significant organizational changes or capital investment.

Engcomp's Analytics team and our associates have extensive experience modeling manufacturing processes, warehouse operations and quantifying the risks associated with large capital projects.

Optimization

What is Optimization? When you think of Optimization, it would be reasonable to describe it as the best outcome given the resources you have available.

“Best” is a relative term, and depending on our objective, it could be lowest cost, maximum profit, least waste, soonest completion time, or highest utilization.

At Engcomp, we work with our clients to develop a model that considers resources and operating constraints but yet still provides a solution that efficiently utilizes these resources and allows the organization to make marked improvements on their performance metrics.

Past optimization projects include transportation schedules, sports field utilization, classroom time tables, locating and rationalizing production facilities, product blending, warehouse space utilization studies, and employee scheduling.