Modern Trends in Trajectory Planning for Two-Tiered Learning Architectures

Abstract

Two-tiered organization provides a compact survey lens for autonomous control (AC): an upper tier supports learning, planning, and high-level (HL) mission reasoning, while a lower tier executes feedback control, safety filtering, and platform-specific constraints. This compact survey reviews hierarchical architectures for artificial intelligence (AI)-driven control with emphasis on mobile, military, and maritime (MMM) information technology (IT) convergence. The paper synthesizes three representative families: classical layered and hierarchical reinforcement learning (HRL) architectures, transformer-based architectures that formulate control as sequence or languageconditioned decision making, and world-model-based (WMB) architectures that use learned predictive dynamics for planning or imagination. The comparison suggests that practical systems in MMM domains should combine verified lower-tier control, data-driven semantic interfaces, predictive world models (WMs), and explicit runtime monitors (RMs).