Reading a hard constraint against coverage
A disqualifying limit is not one objective among many. Spend everything else before you spend it.
Built in·updated 2026-08-03
A hard constraint removes the plan from the ranking entirely. A coverage metric scored linearly costs you a fraction of its weight.
That asymmetry should decide the order you work in: get every hard constraint inside its limit with margin first, then spend what is left on the scored metrics. A plan at 96% coverage that breaches a limit scores nothing; the same plan at 91% coverage that clears it scores most of the points.
Aim inside, not at. Optimizer objectives are targets, not guarantees — the achieved dose lands near them, not on them. Set the objective 5% inside the limit so that normal variation does not cross it.
When coverage genuinely cannot be had. Say so in the write-up. A case where the constraint and the prescription are incompatible is a real finding about the case, and it is more useful to the next planner than a plan that quietly failed.
How to do it in
- Max Dose
- 5 % inside the disqualifying limit
- Weight
- above every target objective roughly 2× the coverage weight
Give the constraint's VOI a Max Dose objective 5 % inside the limit and a weight above every target objective, and leave it there for the whole session.
Precision reports achieved dose against your objectives after each run. Read the constrained structure first, every time. If it is inside with margin, spend the next change on coverage; if it is not, nothing else you do matters.