Neural Information Processing Systems http://nips.cc/

Advancements in large language models (LLMs) have demonstrated their potential in facilitating high-level reasoning, logical reasoning and robotics planning. Recently, LLMs have also been able to generate reward functions for low-level robot actions, effectively bridging the interface between high-level planning and low-level robot control. However, the challenge remains that even with syntactically correct plans, robots can still fail to achieve their intended goals. This failure can be attributed to imperfect plans proposed by LLMs or to unforeseeable environmental circumstances that hinder the execution of planned subtasks due to erroneous assumptions about the state of objects. One way to prevent these challenges is to rely on human-provided step-by-step instructions, limiting the autonomy of robotic systems. Vision Language Models (VLMs) have shown remarkable success in tasks such as visual question answering and image captioning. Leveraging the capabilities of VLMs, we present a novel framework called Robotic Replanning with Perception and Language Models (RePLan) that enables real-time replanning capabilities for long-horizon tasks. This framework utilizes the physical grounding provided by a VLM's understanding of the world's state to adapt robot actions when the initial plan fails to achieve the desired goal. We test our approach within four environments containing seven long-horizion tasks. We find that RePLan enables a robot to successfully adapt to unforeseen obstacles while accomplishing open-ended, long-horizon goals, where baseline models cannot. Find more information at https://replan-lm.github.io/replan.github.io/

Drew McDermott Research on planning for robots is in such a state of flux that there is disagreement about what planning is and whether it is necessary. We can take planning to be the optimization and debugging of a robot's program by reasoning about possible courses of execution. It is necessary to the extent that fragments of robot programs are combined at run time. There are several strands of research in the field; I survey six: (1) attempts to avoid planning; (2) the design of flexible plan notations; (3) theories of time-constrained planning; (4) planning by projecting and repairing faulty plans; (5) motion planning; and (6) the learning of optimal behaviors from reinforcements. More research is needed on formal semantics for robot plans.

The recent advances in computer speed and algorithms for probabilistic inference have led to a resurgence of work on planning under uncertainty. The aim is to design AI planners for environments where there might be incomplete or faulty information, where actions might not always have the same results, and where there might be tradeoffs between the different possible outcomes of a plan. Addressing uncertainty in AI, planning algorithms will greatly increase the range of potential applications, but there is plenty of work to be done before we see practical decision-theoretic planning systems. This article outlines some of the challenges that need to be overcome and surveys some of the recent work in the area. In problems where actions can lead to a number of different possible outcomes, or where the benefits of executing a plan must be weighed against the costs, the framework of decision theory can be used to compare alternative plans.

Planning domains often feature subproblems such as route planning and resource handling. Using static domain analysis techniques, we have been able to identify certain commonly occurring subproblems within planning domains, making it possible to abstract these subproblems from the overall goals of the planner and deploy specialized technology to handle them in a way integrated with the broader planning activities. Although such strategies can be impressive when applied to toy domains, they cannot address highly structured problem domains effectively. However, when knowledge-sparse approaches are supplemented by domain knowledge, they can perform impressively (Bacchus and Kabanza 2000) at the cost of an increased representation burden on the domain designer.

The 1998 Planning Competition at the AI Planning Systems Conference was the first of its kind. Its goal was to create planning domains that a wide variety of planning researchers could agree on to make comparison among planners more meaningful, measure overall progress in the field, and set up a framework for long-term creation of a repository of problems in a standard notation. A rules committee for the competition was created in 1997 and had long discussions on how the contest should go. One result of these discussions was the pddl notation for planning domains. This notation was used to set up a set of planning problems and get a modest problem repository started.

How did TALPLANNER come about? TAL serves as a reference formalism for We use a simple gripper domain as an example. ROBBY only has a single gripper. For many domains, the process is intuitive and straightforward. We imagine that for other domains, the process will be quite complex, and finding a means of automatically generating at least some of the control statements is highly desirable and a challenging research issue.

Because of military drawdowns and the need for additional transportation lift requirements, the United States Marine Corps developed a concept that enabled it to modify a commercial container ship to support deployed aviation units. However, a problem soon emerged in that there were too few people who were expert enough to do the unique type of planning required for this ship. Additionally, once someone did develop some expertise, it was time for him/her to move on, retire, or leave active duty. There needed to be a way to capture this knowledge. Access modules are used to access secondand third-tier mobile facilities that are complexed below decks in support of IMA-level repair capability.

This article reports on an extensive survey and analysis of research work related to machine learning as it applies to automated planning over the past 30 years. Major research contributions are broadly characterized by learning method and then descriptive subcategories. Survey results reveal learning techniques that have extensively been applied and a number that have received scant attention. We extend the survey analysis to suggest promising avenues for future research in learning based on both previous experience and current needs in the planning community. Within the AI research community, machine learning is viewed as a potentially powerful means of endowing an agent with greater autonomy and flexibility, often compensating for the designer's incomplete knowledge of the world that the agent will face and incurring low overhead in terms of human oversight and control.

Planning--the ability to synthesize a course of action to achieve desired goals--is an important part of intelligent agency and has thus received significant attention within AI for more than 30 years. Work on efficient planning algorithms still continues to be a hot topic for research in AI and has led to several exciting developments in the past few years. This article provides a tutorial introduction to all the algorithms and approaches to the planning problem in AI. To fulfill this ambitious objective, I introduce a generalized approach to plan synthesis called refinement planning and show that in its various guises, refinement planning subsumes most of the algorithms that have been, or are being, developed. It is hoped that this unifying overview provides the reader with a brand-name-free appreciation of the essential issues in planning.