Foraging theory is typically used to model animal decision making. We describe an agent such as an autonomous vehicle or software module as a forager searching for tasks. The prey model is used to predict which types of tasks an agent should choose to maximize its rate of reward, and the patch model is used to predict when an agent should leave a patch of tasks and how to choose within-patch search patterns. We expand and apply these concepts to fit an autonomous vehicle control problem and to provide insight into how to make high-level control decisions. We also discuss extensions of the basic models, showing how a risk-sensitive version can be used to alter policies when time or fuel is limited. Throughout the applications, we examine ways an agent can estimate environmental parameters when such parameters are not known.