6 part series
A five-part program for standing up true scheduling optimization in IBM Maximo Application Suite 9 — from why the 7.6 Scheduler was never an optimizer, through the constraint-and-objective model, the labor and location data it lives or dies on, routing and ArcGIS spatial integration, to the Dispatching Dashboard and a phased, Phase-4 rollout.
The 7.6 Scheduler drew what a human decided; Maximo Optimizer decides. This part draws the hard line between graphical scheduling and mathematical optimization — constraint programming and mixed-integer programming — and positions Optimizer honestly as a Phase 4 amplifier that only works on clean data underneath it.
Optimizer's behavior is defined by three things — the constraints it must respect, the objectives you tell it to pursue, and the parameters that bound the solver. This part walks all nine scheduling constraints, the five weighted objective functions, crew scheduling and resource leveling, and the parameters that make or break a run.
Optimizer can only honor constraints that exist as data, and most of that data has to be cleaned before the first run. This is the tactical heart of the series: the prerequisite checklist, the craft/skill and shift cleanup that is the real project, the location coordinates that unlock travel optimization, and a data-readiness checklist to run before you blame the solver.
Travel is where field service wins or loses. This part covers how Optimizer estimates travel — from straight-line distance to real road networks — how route optimization groups and sequences work, and when the optional Esri ArcGIS integration is worth the effort versus straight-line estimation.
The series finale. How an optimized schedule surfaces in Graphical Assignment and the Dispatching Dashboard, the review-adjust-re-optimize rhythm dispatchers actually run, AppPoints licensing math, the Phase 4 rollout timeline, and the three documented use cases — field service, maintenance-window planning, emergency response — with the payoffs that prove Optimizer worked.
