The actorʼs view of automated planning and acting: A position paper

Abstract

Planning is motivated by acting. Most of the existing work on automated planning underestimates the reasoning and deliberation needed for acting; it is instead biased towards path-finding methods in a compactly specified state-transition system. Researchers in this AI field have developed many planners, but very few actors. We believe this is one of the main causes of the relatively low deployment of automated planning applications. In this paper, we advocate a change in focus to actors as the primary topic of investigation. Actors are not mere plan executors: they may use planning and other deliberation tools, before and during acting. This change in focus entails two interconnected principles: a hierarchical structure to integrate the actor's deliberation functions, and continual online planning and reasoning throughout the acting process. In the paper, we discuss open problems and research directions toward that objective in knowledge representations, model acquisition and verification, synthesis and refinement, monitoring, goal reasoning, and integration.

The 2014 International Planning Competition: Progress and Trends

“…In many cases, modeling domains of interest to the broader research community may simply require elicitation from domain experts and formalization within the language constraints of existing tracks. In other cases this may require using more expressivity already available in existing languages or even novel language extensions to include, for example, object fluents (supported by both PDDL 3.1 and RDDL but not used in existing competitions), constraints and timelines (supported by ANML [Smith, Frank, and Cushing 2008]), hierarchical decomposition (Erol, Hendler, and Nau 1994;Shivashankar et al 2013), or more holistic views of the planning system (Ghallab, Nau, and Traverso 2014;Pollack and Horty 1999).…”

Robust plan execution via reconfiguration and replanning

“…While all modalities of a given action achieve the same propositional effects, a given modality models "how" these effects can be achieved. According to the Ghallab et al's terminology [15], modalities are intended to point out the alternative operational model their selection implies. The underlying hierarchical relation defined by the MMA model is an attempt to bridge the gap between the descriptive nature of the PDDL language and the operational representation leading the execution of lower level actions 19 .…”

“…This means that unexpected contingencies may arise at any step of the execution. It is not thus surprising that, as also recently stated in [15], the problem of robust plan execution in real-world domains is a very challenging topic in AI planning, necessary for implementing planning mechanisms in autonomous systems.…”

“…The first two domains are from the Third International Planning Competition 15 ; they have initially been introduced to challenge planners in handling numeric fluents. The third domain has been introduced in our recent works on intelligent supervision of space exploration missions [26,29], while its complete model definition is reported in [39].…”

Intelligent Decision Support Algorithm Based on Self-Adaption Reasoning

“…Intelligent decision support (artificial intelligence planning), also known as automated planning (automated planning), is an important area of artificial intelligence research, and covers a knowledge representation, automatic reasoning, non-monotonic logic, human-computer interaction and cognition science and other areas of cross-disciplinary [6]. The first intelligent decision support research, which originated from the automatic reasoning and knowledge representation, in the 20th century, 90 years ago, has been the use of logical deduction method to be solved, with the focus on classical logic of reasoning.…”