Skip to Main Content
This paper presents arguments for a balanced approach to modeling and reasoning in an autonomous robotic system. The framework discussed uses both declarative and procedural modeling to define the domain, rules, and constraints of the system and also balances the use of deliberative and reactive reasoning during execution. This paper also details the implementations of such an approach on two research rovers and a simulated rover all in a Mars-like environment. Intelligent decision-making capabilities are shown in the context of several unforeseen events, which require action. These events test the system's framework by requiring the system to handle uncertainty in state and resource estimations and in real-world execution. Future work, which further enhances the idea of balanced reasoning, is also discussed.