Scheduling Optimizer

The Scheduling Optimizer automatically works out a schedule for a project (or several projects at once) from its tasks, dependencies, constraints, and available people/equipment/materials — instead of you placing every task on the Gantt chart by hand.

Opening the optimizer

Every project has an Optimizer button in its header. Clicking it opens the Optimizer Vault — a drawer where you run optimizations and review results for that project.

You can also optimize several projects together from a Portfolio’s Strategy tab: select projects in the prioritization table and choose Optimize N Projects to open the same vault scoped to that selection.

Running an optimization

Open the optimizer vault from a project (or a selection of up to 50 projects) and start a run. There is nothing to configure: the run plans every selected project against its priority and desired end date, respecting the shared people, equipment and material, with a fixed search budget (10 seconds, at most 40).

While a run is in progress, the vault shows a solving state; the result loads as a draft — nothing on your real schedule changes until you apply it.

Reading the result

The result is shown in four tabs:

Projekte — one lane per project in the run: the projected end date and whether it is on time or late against the desired end date, the cost of that project’s own resource usage, how many tasks moved compared to the current plan, and the problems the optimizer ran into (e.g. a gate that would be violated, or a task that can’t be resourced), each with a suggestion where possible.

Ressourcen — a utilization heatmap showing how busy each resource (person or equipment) is across the plan, with the existing bookings as context.

Material — what the scheduled tasks still need purchased and by when, derived from their material demand against existing orders and stock, pooled across the whole run. The solver’s own purchase suggestions are counted in the run total and are filed as purchase requests when you apply, each with an arrival day and an order-by day. A good needed by several projects is requested once, by the first project applied.

Zeitplan — the draft rendered as a Gantt chart, with the current (live) plan shown as a ghost overlay so you can see exactly what moved. Gates, milestones and events appear as markers, each shown as held, missed, unknown, or already finished.

Cost per project covers that project’s resource usage; the total across the run adds the suggested purchase orders, shown separately.

If the optimizer can’t find a workable plan at all, the vault shows a diagnosis instead of a schedule, to help you see what’s blocking it (e.g. missing resources or conflicting constraints).

Pinning tasks and occasions

Within the draft schedule, you can pin a task to lock in its computed start time while you keep exploring other changes. Pinning is provisional — it only affects the current draft. If you want a pinned position to survive future optimization runs, you can promote it: a task is written back as a real “Not before” constraint (see Segments, Task Packages, Tasks & Subtasks), while a gate or milestone is written back as a fixed date (both bounds). Promoted occasions can be un-pinned again from the occasion list.

An event with a booked slot keeps that slot and is never moved by the optimizer. A placeholder event without a slot is positioned by the plan, like a milestone.

Applying a draft

When you’re happy with a draft, Apply commits it: task dates and resource bookings are written to the real schedule. Applying is also how an early-phase project (Initial, Definition, or Planning) enters Execution — running and applying the optimizer is required to move a project into Execution, and the project’s phase advances automatically once you apply.

A draft is applied once per project; applying the same parts again requires a new run. Projects that are cancelled or closed cannot receive an optimization. Applying and rolling back require an organisation admin role.

If you later move a project that has an applied optimization back out of Execution (for example, back to Planning), GreenVee warns you first, then restores the schedule to what it was before the optimizer was applied: task positions, assignees and dates go back to their pre-apply values where nobody has changed them since, the optimizer’s bookings and open purchase requests are retired, and a cost baseline the apply created is released. Work already logged is not touched.

For Project Managers

Run the optimizer before committing to an execution start. Review all four tabs — nothing is final until you Apply.