TABLE 7.

Characteristics of reinforcement learning based boundary tracking.

Characteristic	Description
Objective	Track boundaries of phenomenon
State Representation	Environmental features, object boundaries
Action Space	Movement actions, sensing actions
Reward Function	Rewards for accurate boundary tracking, penalties for errors
Observation	Sensor data for measuring phenomenon (e.g., vision sensor or intensity/concentration measurements)
Challenges	Noisy or ambiguous boundaries, initial search phase
Citations	Julian and Kochenderfer, (2019); Viseras et al. (2021)