Applied AI for operations

Risk-informed intervention planning for constrained transit operations.

This page turns the atlas into an action surface. It uses a simple predictive missed-connection model and a budget-constrained optimiser to recommend where a small set of interventions should go first.

Overview Atlas Operations Benchmark Research Review
Predictive risk scoring Budgeted optimisation Operational decision support

Recommended actions

Where a limited support budget should go first

Loading

Baseline exposure

Origins ranked by reliable access before intervention

Origin Municipality Population weight Reliable access Top destinations

Method

What the optimiser is doing

The score discounts weighted opportunities by predicted missed-connection probability, then chooses one intervention per origin under a fixed budget. The current payload is loaded from ./data/operations_plan.json.