APS performs well within the boundaries it was designed for, but three limitations become visible in more complex environments. First, APS relies on rules and parameters that planners must configure and keep accurate — in factories where products or processes change frequently, maintaining this master data is a significant ongoing effort. Second, APS heuristics explore only a fraction of all possible schedules, so in high-mix environments with sequence-dependent setups, the gap between what APS finds and the true optimum can be substantial. Third, most APS systems require planners to manually trigger replanning when disruptions occur, they don't automatically minimise the ripple effect across the rest of the schedule.
