Tag
NASA's Perseverance rover set a record for the farthest distance driven on another world (45.22 km), enabled by autonomous navigation upgrades including Global Localization, radiation-tolerant software techniques, and generative AI path planning.
本文提出一个可观察证据标准，用于评估工业AI规划系统中新增能力是否真正带来运营价值，并通过Favorita、M5和Walmart数据的回放实验揭示单一指标评估的三大局限。
Chunk is a time-blocking app for macOS that combines tasks and calendar in one place, supporting drag-to-block, two-way sync with Google/Outlook/Apple Calendar, local-only data storage via iCloud, and AI week planning powered by Claude.
This paper introduces an approach for automatically generating generalized plans in Lean with formal proofs of completeness using LLMs, evaluated on benchmark domains showing significant advancements in automatic plan verification.
Opus 5.5 demonstrated extended planning capabilities by spending over 20 minutes on complex puzzles, solving ones previously only solved by Astra.
This article compares four AI planning tools—OpenSpec, Spec Kit, BMAD, and Kiro—by measuring the manual actions needed to plan a simple app from a one-sentence idea, revealing large differences in efficiency and features.
This paper introduces a novel approach using large language models to automatically generate instance-generation programs for PDDL planning domains, ensuring soundness and diversity.
This paper adapts two plan deordering techniques from classical planning to hierarchical task network planning, showing a substantial reduction in ordering constraints while maintaining plan validity.
This paper evaluates open-weight large language models for PDDL domain repair in AI planning, showing they outperform symbolic baselines but struggle with reliable test constraint satisfaction.
/wayfinder is a planning tool that helps developers map out ambitious projects by figuring out actionable steps and maintaining a map in an issue tracker like GitHub or Linear. It's described as revolutionary for planning coding work.
Google DeepMind is partnering with the UK government to develop an AI-powered planning prototype using Gemini, aiming to halve the time for processing householder planning applications and help build 1.5 million new homes by 2029.
Introducing /improve, a tool that uses Claude's most capable model to audit your codebase, identify bugs, performance issues, tech debt, missing tests, and write plans that cheaper models can execute.
This paper introduces the concept and architecture of a Business World Model (BWM), a specialized world model for business environments that encodes states, dynamics, constraints, and objectives to support autonomous decision-making and goal-driven planning.
This paper introduces OCLGen, a compute-efficient test-time search algorithm that integrates generative planning models with a classical Open-Closed List framework, improving solution quality across combinatorial planning domains.
This paper investigates how to encode factored planning tasks (FTS) into SAT, proposing multiple encoding strategies and analyzing the impact of task transformations on SAT-based planning performance. It aims to extend SAT solving to more compact planning representations beyond heuristic search.