Planning maintenance around the trains

Algorithm Lab · Open, no account

How CP-SAT works

Constraint programming turns work into choices and rules. SAT search rules out impossible combinations and helps find the best valid schedule.

1. 01

   Model the work

   Define when each job can run, the rules it must respect and the cost of delay.

2. 02

   Search within the rules

   Eliminate impossible choices, explore alternatives and learn from conflicts.

3. 03

   Prove and explain

   Compare the best schedule with a proven bound, then check the result independently.

Real OR-Tools solves. Six example jobs. Change the rules and see why work moves.

NebulaX PS1 · Open, no account

Railway track access scheduler

Assigns 54 contracted activities to weeks across two lines, checks its own answer against the nine hard rules, scores all three scenarios and writes the submission files. Load the published instance or upload your own.

No sign-in required. The scheduler checks complete plans before display and export.

RailPlan workspace · Account required

Overnight engineering planning

The wider prototype: contractors submit maintenance requests, a planner reviews them, schedules the night against the constraint engine and publishes the agreed slots.

For provisioned planner and contractor accounts. Roles are assigned by a workspace administrator, so signing up is not part of this prototype.