A Unified Framework for Real-Time Failure Handling in Robotics Using Vision-Language Models, Reactive Planner and Behavior Trees
(2025) 21st IEEE International Conference on Automation Science and Engineering, CASE 2025 p.887-894- Abstract
Robotic systems often face execution failures due to unexpected obstacles, sensor errors, or environmental changes. Traditional failure recovery methods rely on predefined strategies or human intervention, making them less adaptable. This paper presents a unified failure recovery framework that combines Vision-Language Models (VLMs), a reactive planner, and Behavior Trees (BTs) to enable real-time failure handling. Our approach includes pre-execution verification, which checks for potential failures before execution, and reactive failure handling, which detects and corrects failures during execution by verifying existing BT conditions, adding missing preconditions and, when necessary, generating new skills. The framework uses a scene... (More)
Robotic systems often face execution failures due to unexpected obstacles, sensor errors, or environmental changes. Traditional failure recovery methods rely on predefined strategies or human intervention, making them less adaptable. This paper presents a unified failure recovery framework that combines Vision-Language Models (VLMs), a reactive planner, and Behavior Trees (BTs) to enable real-time failure handling. Our approach includes pre-execution verification, which checks for potential failures before execution, and reactive failure handling, which detects and corrects failures during execution by verifying existing BT conditions, adding missing preconditions and, when necessary, generating new skills. The framework uses a scene graph for structured environmental perception and an execution history for continuous monitoring, enabling context-aware and adaptive failure handling. We evaluate our framework through real-world experiments with an ABB YuMi robot on tasks like peg insertion, object sorting, and drawer placement, as well as in AI2-THOR simulator. Compared to using pre-execution and reactive methods separately, our approach achieves higher task success rates and greater adaptability. Ablation studies highlight the importance of VLM-based reasoning, structured scene representation, and execution history tracking for effective failure recovery in robotics.
(Less)
- author
- Ahmad, Faseeh
LU
; Ismail, Hashim
LU
; Styrud, Jonathan
; Stenmark, Maj
LU
and Krueger, Volker
LU
- organization
-
- Robotics and Semantic Systems
- LTH Profile Area: AI and Digitalization
- ELLIIT: the Linköping-Lund initiative on IT and mobile communication
- NEXTG2COM – a Vinnova Competence Centre in Advanced Digitalisation
- LU Profile Area: Natural and Artificial Cognition
- LTH Profile Area: Engineering Health
- Department of Computer Science
- publishing date
- 2025
- type
- Chapter in Book/Report/Conference proceeding
- publication status
- published
- subject
- host publication
- 2025 IEEE 21st International Conference on Automation Science and Engineering, CASE 2025
- pages
- 8 pages
- publisher
- IEEE Computer Society
- conference name
- 21st IEEE International Conference on Automation Science and Engineering, CASE 2025
- conference location
- Los Angeles, United States
- conference dates
- 2025-08-17 - 2025-08-21
- external identifiers
-
- scopus:105018322736
- ISBN
- 9798331522469
- DOI
- 10.1109/CASE58245.2025.11164021
- language
- English
- LU publication?
- yes
- id
- 2f6d895f-afd2-431e-a284-55a53a4162bc
- date added to LUP
- 2026-01-08 15:54:01
- date last changed
- 2026-01-08 15:54:28
@inproceedings{2f6d895f-afd2-431e-a284-55a53a4162bc,
abstract = {{<p>Robotic systems often face execution failures due to unexpected obstacles, sensor errors, or environmental changes. Traditional failure recovery methods rely on predefined strategies or human intervention, making them less adaptable. This paper presents a unified failure recovery framework that combines Vision-Language Models (VLMs), a reactive planner, and Behavior Trees (BTs) to enable real-time failure handling. Our approach includes pre-execution verification, which checks for potential failures before execution, and reactive failure handling, which detects and corrects failures during execution by verifying existing BT conditions, adding missing preconditions and, when necessary, generating new skills. The framework uses a scene graph for structured environmental perception and an execution history for continuous monitoring, enabling context-aware and adaptive failure handling. We evaluate our framework through real-world experiments with an ABB YuMi robot on tasks like peg insertion, object sorting, and drawer placement, as well as in AI2-THOR simulator. Compared to using pre-execution and reactive methods separately, our approach achieves higher task success rates and greater adaptability. Ablation studies highlight the importance of VLM-based reasoning, structured scene representation, and execution history tracking for effective failure recovery in robotics.</p>}},
author = {{Ahmad, Faseeh and Ismail, Hashim and Styrud, Jonathan and Stenmark, Maj and Krueger, Volker}},
booktitle = {{2025 IEEE 21st International Conference on Automation Science and Engineering, CASE 2025}},
isbn = {{9798331522469}},
language = {{eng}},
pages = {{887--894}},
publisher = {{IEEE Computer Society}},
title = {{A Unified Framework for Real-Time Failure Handling in Robotics Using Vision-Language Models, Reactive Planner and Behavior Trees}},
url = {{http://dx.doi.org/10.1109/CASE58245.2025.11164021}},
doi = {{10.1109/CASE58245.2025.11164021}},
year = {{2025}},
}