# Capacity forecast

## What is this?

The capacity forecast (Plan → Capacity forecast) shows per station type
whether your stations cover the scheduled and not-yet-scheduled tests of the
coming weeks — for example "are our climate chambers enough next quarter?"
or "do we need more cyclers?". It uses the same schedule as the occupancy
calendar and the double-booking warning, and shows weekly utilisation, the
highest number of concurrent tests and how many stations are missing.
What-if scenarios let you try additional stations or a moved project without
changing the plan.

## How to use it

- **Horizon and target utilisation:** At the top, choose 8, 13 or 26 weeks
  and a target utilisation of 70, 80 or 90 %. The target is the buffer for
  retests, failures and changeovers — 80 % is a good default. The first week
  starts now, not last Monday.
- **One card per station type:** It shows the number of stations ready for
  use, the overall utilisation for the horizon and an assessment:
  - **Capacity is sufficient** — no week above target utilisation and never
    more concurrent tests than stations.
  - **Above target utilisation** — little buffer in some weeks.
  - **Peak load** — at times more tests are meant to run at once than there
    are stations, even if weekly utilisation looks harmless. Staggering
    starts is often enough.
  - **Not enough capacity** — plan and backlog don't fit under the target
    within the horizon; the card says how many stations are missing.
- **Weekly bars:** Utilisation per week in percent. Blue = normal, orange =
  more concurrent tests than stations (shown below, e.g. "3/2"), red = above
  target utilisation. The dashed line is the target. Hover over a bar to see
  booked and available hours.
- **Backlog:** Open test cases without a place in the plan — no station, no
  computable date (sequence without a start) or a date that has already
  passed. The card shows the hours and by when the backlog fits into free
  capacity. Without an assigned station, the type is derived from the
  allowed stations or station types of the test case template when it is
  unique.
- **What if — additional stations:** Use "+"/"−" on a card to calculate with
  hypothetical extra stations of that type.
- **What if — move a project:** Pick a project at the top and enter the
  weeks (negative = earlier). All of the project's bookings are moved in the
  calculation.
- **Reset scenario:** Removes all what-if assumptions. Scenarios are never
  saved; they only live in the page address and can be shared as a link.
- **Via agent/MCP:** The `get_capacity_forecast` tool returns the same
  figures, including what-if scenarios (see [Agent tokens](https://www.testlabiq.com/en/help/agent-access)).

## How it is calculated

- **Available hours:** Every station counts around the clock (24/7), minus
  its next maintenance window and its reservations. Stations with the status
  "Out of service" are not counted.
- **Booked hours:** Open test cases (pending or in progress) with a station,
  in their computed time span from the schedule. Attended tests only occupy
  their station during working hours.
- **Missing stations:** (booked hours + backlog − available hours × target)
  ÷ (hours in horizon × target), rounded up.
- As with the occupancy calendar and conflict check, the 50 most recent
  projects are included.

## Common issues

- **"A type shows 0 % utilisation although tests are planned there."** The
  test cases have no station assigned or their sequence has no start — they
  then count as backlog, not as bookings. Assign a station or set the
  sequence start.
- **"Test cases without a unique station type are not included."** The test
  case has no station and its template allows no station type or several.
  Assign a station, or set exactly one allowed type (or stations of one
  type) on the template.
- **"Utilisation looks too low."** A station is available 24/7. For purely
  attended benches with 40 working hours per week, the achievable
  utilisation is therefore around 25 %. Watch the peak load there instead.
- **"A station in maintenance is still counted."** Only the status "Out of
  service" removes a station completely. Maintenance takes effect through
  the maintenance window you maintain (see [Stations](https://www.testlabiq.com/en/help/stations)).
- **"I can't select the project to move."** Only projects with bookings in
  the selected horizon are offered.
- **"Notice: very large plans are truncated."** More than 50 active
  projects, 200 stations or 1,000 reservations in the horizon — the
  forecast then works with a subset.

---

Source: https://www.testlabiq.com/en/help/capacity-forecast