Using logic to handle conflicts between system, component, and infrastructure goals in complex robotic architectures

Abstract

Complex robots with many interacting components in their control architectures are subject to component failures from which neither the control architecture nor the implementing infrastructure can recover. Moreover, the operating conditions for these components might be at odds with goals the robot might have adopted (e.g., through external commands or in the course of the execution of the current task).We argue that the best (if not the only) way to resolve any difficulties that arise from the different requirements at the agent, component and infrastructure levels is to use a common formal logical goal representation for all three layers. We discuss how these representations can be integrated into a complex robotic architecture and demonstrate in an experimental evaluation on a robot how the architecture can recover from a failure situation that it would not have been able to handle without explicit multi-level unified goal representations and their associated monitoring and reasoning processes.

Crossing Boundaries: Multi-Level Introspection in a Complex Robotic Architecture for Automatic Performance Improvements

“…Agent-level introspection in DIARC is performed by the DIARC goal manager, which selects the best available action to achieve the agent's goals in its current state (Schermerhorn and Scheutz 2010). The goal AT(ROBOT,RESEARCH LAB) and the actions working toward it are monitored to ensure progress, and new actions are selected based on agent and world states.…”

Abstract planning for reactive robots

“…The FODD-PLANNER was integrated into the DIARC robot architecture, which uses the Java-based Agent Development Environment (ADE) [14] as a framework for implementing architectural components (e.g., perceptual processing, navigation, action planning and natural language processing). Such an integration presents challenges at multiple levels.…”

“…State updates and goal requests received from the robot architecture are submitted to the planner process and actions returned from policy consultation are routed back to the goal manager by the ADE planner component. Communication between architectural components using predicates is well-established in ADE [14], so at the syntactic level, the interface is straightforward once the vocabulary is established. The search and report task was captured using five deterministic actions and one sensing action: movehall from one location to another, enter a room, exit a room, report a red box, deliver at a certain location, and the sensing action lookfor as described above.…”

“…Concretely, we show how the FODD-PLANNER [8] can be integrated into the DIARC robotic architecture [14] and solve tasks requiring high-level goals demonstrated in two application domains: a search and report task capturing a rescue operation [16] and a search and manipulation task where various manipulation activities related to rescue operations have to be performed [3].…”

Argumentation-Based Online Incremental Learning

“…Therefore, it is important to note that confronting unforeseen failures is mostly the default state for GPSRs, rather than an exceptional state as often described in the literature. There are some solutions for external failure recovery in the literature, which involve using simulations for the prediction of external failures [5] and logic-based reasoning to account for external failures [6], [7]. However, in most of these cases, the solutions are proposed for specific applications.…”