Progress control and S-curve

Construction progress control with an S-curve and a schedule forecast that admits uncertainty

Instead of telling you "we finish on the 30th", the forecast simulates thousands of scenarios over your schedule and gives you a range: the date at 50%, 80% and 95% probability.

The problem

The delivery date nobody dares to move until it is too late

The schedule was built at the start, progress is reported as an overall percentage and the real delay shows up when there is no way left to recover it.

A single date hides the risk. If a critical path activity slips two weeks, how likely is it that the contract deadline is missed?

How it works

From the schedule to the deadline risk

  1. You load the schedule

    You build it on the platform or import it from another scheduling program: MSPDI XML and, with the conversion service active, .mpp, .xer or .pmxml files. It comes in as a variant you can compare with the baseline.

  2. You record progress

    Each task's percentage and the actual cost feed the earned value: CPI, SPI and the final cost projection. If there is cost with no active line item, "—" is shown.

  3. You look at the S-curve

    The planned curve is derived from the dates of the work packages. Earned value and actual cost are shown as today's point.

  4. You check the forecast

    A Monte Carlo simulation propagates delays through the dependencies and returns the projected finish, P50, P80, P95 and the tasks that turn out critical most often.

What sets it apart

What you can check, not just take on faith

A range, not a date
It shows P50, P80 and P95. "At risk" means the P80 falls after the committed date; if you didn't load that date, it says it can't know.
The assumption is in plain view
Each task draws a factor from 0.9 to 1.5 over its remaining work. It is an assumption, not a measurement of your project, and the screen declares it.
The critical tasks that matter
It lists up to five tasks that are critical in at least 20% of the scenarios.
Reproducible
The simulation uses a fixed seed: with the same data it gives the same result.

What it doesn't do

The limits, stated up front

  • The forecast draws each task independently, so it probably underestimates the real dispersion. It is a reference, not a guarantee.
  • The variation factors (0.9, 1.0 and 1.5) are assumptions; they are not calibrated with your project's history.
  • Earned value and actual cost are a single point for today; there is no historical curve of the actual.
  • Work packages without dates are left out of the curve, and the screen counts how many there are.
  • Importing .mpp, .xer or .pmxml depends on a conversion service that must be enabled; if it isn't, it warns with an error and does not deliver an empty schedule.

Compared with what you use today

Compared with the Excel schedule and the planning program

Spreadsheets
A hand-made bar chart doesn't propagate delays through dependencies. Here it does, and it measures the risk.
Desktop planning programs
They give the critical path, rarely a probability of meeting the deadline. Here you import from them and add the range.
Generic ERP
It doesn't model activities with dependencies or construction earned value.

Frequently asked questions

What people usually ask before trying it

What do I need for the forecast?

A schedule with tasks, dates, dependencies and percent complete. If there are circular dependencies or everything is at 100%, the screen says so instead of calculating.

Is it an exact prediction?

No. It is a simulation with declared assumptions; it helps you decide whether to act, not to promise a date.

What is P80?

The date that is met in 80% of the simulated scenarios. If it falls after the contract deadline, the project shows as at risk.

Next step

Let's load your schedule and look at the deadline risk

Tell us about your project and we'll show you this module with your own data, no commitment.

S-curve and progress control software · Renka Consulting