Abstract: Systems and methods for handling autonomous vehicle faults are provided. A system includes a plurality of function nodes arranged in a directed graph architecture. The function nodes include a plurality of detector nodes and a plurality of fault handler nodes. Each detector node is communicatively connected to at least one function node and an associated fault handler node. The detector node is configured to obtain output from the function node, detect a fault, and provide a fault event to the associated fault handler node. Each fault handler node can be associated with a fault severity and a corresponding vehicle action. The associated fault handler node can receive the fault event from the detector node and initiate a fault response. The fault response can include initiating a respective vehicle action in response to the fault.

Abstract: Systems and methods for managing a fleet of vehicles are provided. In one example embodiment, a computer-implemented method includes obtaining data representing vehicle state information associated with one or more vehicles among a fleet of vehicles at one or more times. The method includes predicting a future vehicle state associated with the one or more vehicles at one or more future times. The method includes scheduling a vehicle assignment in a predetermined set of vehicle assignments for a selected vehicle among the one or more vehicles before a first future time among the one or more future times. The method includes transmitting a command signal to a computing system associated with the selected vehicle based at least in part on the scheduled vehicle assignment.