Evidence-Gated Task and Motion Planning with Vision-Language Models

APIVOT: Adaptive Planning with Interleaved Vision-Language Thoughts

This work proposes APIVOT, a VLM-based planner that adaptively interleaves language and visual thoughts for long-horizon planning that outperforms general-purpose VLMs and prior planning frameworks, achieving the largest gains in spatially constrained settings.

Large Language Model-Driven Symbolic Planning for Long-Horizon Robotic Manipulation Tasks

Cross-modal foundation models are increasingly used for robotic task understanding and planning. However, connecting multimodal observations and natural language instructions to symbolic planners and executable robot actions remains a key challenge in embodied AI. In addressing this challenge, existing methods often rely on structured inputs and lack robustness to language ambiguity, with limited real-world validation. We propose VLA-SP (Vision-Language-Action via Symbolic Planning), a two-stage Embodied Vision-Language-Action framework. VLA-SP consists of a Multi-modal Information Processing Module and a PDDL-based Vision-Language-Action Module, enabling fully automated robotic execution from speech and vision inputs. First, our framework detects task-relevant objects using YOLO and transcribes voice instructions with Faster-Whisper. After obtaining the visual observations and language instructions, the LLM fuses these multimodal inputs and converts them into a planner-verifiable problem.pddl instance under a reusable domain.pddl. Then, Fast Downward validates the generated instance and, upon successful verification, produces a multi-step symbolic plan. If any issues are detected during verification, planner feedback is used for traceable failure diagnosis and iterative self-correction. Once a valid symbolic plan is obtained, the symbolic actions are further grounded into 6-DoF poses and executed through predefined robot skills. Finally, we conduct extensive experiments across seven planning domains in simulation and six real-world manipulation tasks on two robotic platforms, X-Arm and Franka. The experimental results demonstrate that VLA-SP enables reliable execution with high success rates across diverse manipulation tasks. In particular, the complete framework achieves an average success rate of 82% across the six real-world tasks. Results from both simulated and real-world experiments fur-ther demonstrate the strong interpretability, executability, and cross-platform applicability of our framework. Further details are available here.

Predictive vision-language monitoring for proactive safety in robot task execution

Robots that execute language-conditioned tasks in dynamic environments often rely on feedback only after an action has failed, which can be insufficient when failures involve collisions or workspace conflicts. This paper presents a predictive monitoring framework that uses Vision-Language Models (VLMs) to assess near-future execution risk during robot task execution. The framework first generates structured plans with action execution conditions and a plan-level fallback action. During execution, a monitoring module combines visual observations, the current action, and the relevant execution conditions to estimate whether a condition is likely to be violated within a short future time window. When the predicted risk exceeds a task-specific threshold, the robot halts the current action, executes the fallback behavior, and replans from the updated state. We evaluate the approach in Gazebo simulation on mobile navigation with a moving human obstacle and manipulation with two robot arms sharing a workspace. Across controlled collision-risk settings, the proposed method achieves higher task success rates than reactive VLM-based baselines while requiring fewer replanning events than conservative current-state precondition checking. The results indicate that predictive vision-language monitoring can improve task completion in simulated dynamic robot tasks, while remaining subject to limitations such as VLM latency, prompt sensitivity, and evaluation beyond simulation.

MulPlanLM: multimodal robotic task planning with vision-language models and physical feedback

StellaVLA: In-Context Structured Demonstration for Generalizable Vision-Language-Action Models

Vision-Language-Action (VLA) models can follow instructions and manipulate objects, but their performance often collapses out of distribution (OOD), when the scene, viewpoint, or object differs from training. Adapting to each new situation typically requires collecting more data and fine-tuning. We present StellaVLA, a framework that instead adapts at test time by conditioning on a single retrieved demonstration. The key idea is to move beyond imitating what an expert did and instead convey why: an automated offline pipeline converts each raw trajectory into a structured demonstration, e.g., a task plan, sub-goal descriptions, and verbalized 3D motion, at zero human-annotation cost. Provided as in-context guidance, this structured demonstration lets the policy reason about the task rather than mimic a pixel trajectory, which also makes it transferable across embodiments (real-robot, human-hand, or XR demonstrations). A parallel dual-training design internalizes this reasoning during training through a joint action-and-language objective, while inference uses the action expert alone, preserving real-time, high-frequency control with no added latency. On the VLA-Arena leaderboard(Aug 1, 2026), StellaVLA ranks first with an overall score of 0.63, versus 0.44 and 0.22 for the strong prior models ($\pi_{0.5}$ and LingBot-VLA), and it further leads on LIBERO with 98.8% average success rate and LIBERO-Plus with 85.1% success rate. Our real-robot benchmark demonstrates that StellaVLA can use both human/robot demos and human-to-robot (XR) demos as in-context structured demonstration to help VLA model adapt to OOD tasks.

Cortex: A Bidirectionally Aligned Embodied Agent Framework for Long-horizon Manipulation

Cortex is introduced, a bidirectionally aligned embodied agent framework with a customized planning interface that conveys executable and tractable subtask plans from high-level VLM to low-level VLA and enables zero-shot completion of unseen real-world long-horizon tasks.