# Failure Detection for Surgical Robot Imitation Policies via Flow-Matching World Modeling

arXiv: https://arxiv.org/abs/2607.27511 · HTML: https://mlsanity.com/paper/2607.27511/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Medical and surgical world models, one of 11 directions in World models.
- Builds on 14 of its 31 references that are in the map.
- First posted on arXiv 29 Jul 2026.

## Placed in

- World models / [Medical and surgical world models](https://mlsanity.com/wm/medical-and-surgical-world-models/)
- Diffusion models / [Diffusion and flow policies for robotics](https://mlsanity.com/diffusion/diffusion-and-flow-policies-for-robotics/)
- Offline RL and imitation learning from data / [Language-conditioned policies and VLAs](https://mlsanity.com/offlinerl/language-conditioned-policies-and-vlas/)

## Builds on (most cited first, 14 of 14)

- [Flow Matching for Generative Modeling](https://mlsanity.com/paper/2210.02747/) (2022)
- [DINO-WM: World Models on Pre-trained Visual Features enable Zero-shot Planning](https://mlsanity.com/paper/2411.04983/) (2024)
- [SAFE: Multitask Failure Detection for Vision-Language-Action Models](https://mlsanity.com/paper/2506.09937/) (2025)
- [Back to the Features: DINO as a Foundation for Video World Models](https://mlsanity.com/paper/2507.19468/) (2025)
- [How Far Are Surgeons from Surgical World Models? A Pilot Study on Zero-shot Surgical Video Generation with Expert Assessment](https://mlsanity.com/paper/2511.01775/) (2025)
- [Cosmos-H-Surgical: Learning Surgical Robot Policies from Videos via World Modeling](https://mlsanity.com/paper/2512.23162/) (2025)
- [Foundational World Models Accurately Detect Bimanual Manipulator Failures](https://mlsanity.com/paper/2603.06987/) (2026)
- [Surgical Vision World Model](https://mlsanity.com/paper/2503.02904/) (2025)
- [World Models for General Surgical Grasping](https://mlsanity.com/paper/2405.17940/) (2024)
- [RC-NF: Robot-Conditioned Normalizing Flow for Real-Time Anomaly Detection in Robotic Manipulation](https://mlsanity.com/paper/2603.11106/) (2026)
- [Towards Suturing World Models: Learning Predictive Models for Robotic Surgical Tasks](https://mlsanity.com/paper/2503.12531/) (2025)
- [SutureBot: A Precision Framework & Benchmark For Autonomous End-to-End Suturing](https://mlsanity.com/paper/2510.20965/) (2025)
- [VFMF: World Modeling by Forecasting Vision Foundation Model Features](https://mlsanity.com/paper/2512.11225/) (2025)
- [Cosmos-Surg-dVRK: World Foundation Model-based Automated Online Evaluation of Surgical Robot Policy Learning](https://mlsanity.com/paper/2510.16240/) (2025)

## Built on by (most cited first, 3 of 3)

- [World Models for Embodied Intelligence: From Plausible to Controllable to Actionable](https://mlsanity.com/paper/2609.16697/) (2026)
- [DreamLedger: Where to Refuse World-Model Imagination Using Execution-Settled Credit](https://mlsanity.com/paper/2608.23863/) (2026)
- [FARM: Reading Failure Signals from the Internal Predictive States of a Frozen Robotic World Model](https://mlsanity.com/paper/2609.11445/) (2026)

## Abstract (arXiv)

Imitation learning has shown increasing promise for autonomous robotic surgery, yet safe deployment remains challenging due to the safety-critical nature of surgical tasks and the complexity and variability of surgical environments. Failure detection is therefore an essential safeguard, but its development remains difficult due to the challenges of scarce failure data, highly variable manipulation dynamics, and the need to balance missed detections against disruptive false alarms. To address these challenges, we introduce FoMo-FD (Flow-Matching World Model for Failure Detection), a failure detection method that learns nominal short-horizon visual dynamics with an action-conditioned flow-matching world model. FoMo-FD scores the inverse-transport nonconformity of observed endpoint latents, enabling window-level detection of visual-action inconsistencies without requiring failure demonstrations. Detection thresholds are obtained by conformal calibration on successful executions, yielding task-specific alarms without assuming future failure types. We evaluate FoMo-FD on four surgically relevant manipulation tasks with twenty failure modes across simulation and real-world experiments using the da Vinci Research Kit (dVRK). Results show that FoMo-FD outperforms observation-level anomaly baselines and a prediction-error variant of the same world model, with the wrist-camera view achieving the strongest performance, including a 96.6% failure detection rate (FDR) at a 1.3% false alarm rate (FAR).