# Safe-Night VLA: Seeing the Unseen via Thermal-Perceptive Vision-Language-Action Models for Safety-Critical Manipulation

arXiv: https://arxiv.org/abs/2603.05754 · HTML: https://mlsanity.com/paper/2603.05754/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Placed in Safety and adversarial attacks, one of 13 directions in Vision-language-action models.
- Builds on 25 of its 37 references that are in the map.
- First posted on arXiv 5 Mar 2026.

## Placed in

- Vision-language-action models / [Safety and adversarial attacks](https://mlsanity.com/vla/safety-and-adversarial-attacks/)

## Builds on (most cited first, 25 of 25)

- [Scalable Diffusion Models with Transformers](https://mlsanity.com/paper/2212.09748/) (2022)
- [RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control](https://mlsanity.com/paper/2307.15818/) (2023)
- [OpenVLA: An Open-Source Vision-Language-Action Model](https://mlsanity.com/paper/2406.09246/) (2024)
- [Octo: An Open-Source Generalist Robot Policy](https://mlsanity.com/paper/2405.12213/) (2024)
- [GR00T N1: An Open Foundation Model for Generalist Humanoid Robots](https://mlsanity.com/paper/2503.14734/) (2025)
- [3D-VLA: A 3D Vision-Language-Action Generative World Model](https://mlsanity.com/paper/2403.09631/) (2024)
- [π*0.6: a VLA That Learns From Experience](https://mlsanity.com/paper/2511.14759/) (2025)
- [Control Barrier Function Based Quadratic Programs for Safety Critical Systems](https://mlsanity.com/paper/1609.06408/) (2016)
- [RoboAgent: Generalization and Efficiency in Robot Manipulation via Semantic Augmentations and Action Chunking](https://mlsanity.com/paper/2309.01918/) (2023)
- [ContactDB: Analyzing and Predicting Grasp Contact via Thermal Imaging](https://mlsanity.com/paper/1904.06830/) (2019)
- [Visual Prompt Multi-Modal Tracking](https://mlsanity.com/paper/2303.10826/) (2023)
- [SafeDiffuser: Safe Planning with Diffusion Probabilistic Models](https://mlsanity.com/paper/2306.00148/) (2023)
- [SafeVLA: Towards Safety Alignment of Vision-Language-Action Model via Constrained Learning](https://mlsanity.com/paper/2503.03480/) (2025)
- [Eagle 2.5: Boosting Long-Context Post-Training for Frontier Vision-Language Models](https://mlsanity.com/paper/2504.15271/) (2025)
- [OmniVLA: An Omni-Modal Vision-Language-Action Model for Robot Navigation](https://mlsanity.com/paper/2509.19480/) (2025)
- [Robotic Perception of Transparent Objects: A Review](https://mlsanity.com/paper/2304.00157/) (2023)
- [ReVLA: Reverting Visual Domain Limitation of Robotic Foundation Models](https://mlsanity.com/paper/2409.15250/) (2024)
- [STARE-VLA: Progressive Stage-Aware Reinforcement for Fine-Tuning Vision-Language-Action Models](https://mlsanity.com/paper/2512.05107/) (2025)
- [Audio-Visual Efficient Conformer for Robust Speech Recognition](https://mlsanity.com/paper/2301.01456/) (2023)
- [ShapeClipper: Scalable 3D Shape Learning from Single-View Images via Geometric and CLIP-based Consistency](https://mlsanity.com/paper/2304.06247/) (2023)
- [TRansPose: Large-Scale Multispectral Dataset for Transparent Object](https://mlsanity.com/paper/2307.05016/) (2023)
- [Toward a Holistic Evaluation of Robustness in CLIP Models](https://mlsanity.com/paper/2410.01534/) (2024)
- [SafeFlow: Safe Robot Motion Planning with Flow Matching via Control Barrier Functions](https://mlsanity.com/paper/2504.08661/) (2025)
- [OmniVLA: Physically-Grounded Multimodal VLA with Unified Multi-Sensor Perception for Robotic Manipulation](https://mlsanity.com/paper/2511.01210/) (2025)
- [UniConFlow: A Unified Constrained Flow-Matching Framework for Certified Motion Planning](https://mlsanity.com/paper/2506.02955/) (2025)

## Built on by (most cited first, 5 of 5)

- [Vision-Language-Action Safety: Threats, Challenges, Evaluations, and Mechanisms](https://mlsanity.com/paper/2604.23775/) (2026)
- [SafeDojo: Safe Reinforcement Learning for VLA via Interactive World Model](https://mlsanity.com/paper/2606.20698/) (2026)
- [Security of Foundation-Model-Powered Embodied Agents: Attack Surfaces, Attacks, Defenses, and Evaluation](https://mlsanity.com/paper/2608.16843/) (2026)
- [Security and Privacy in Large-Model-Driven Embodied Agents: Attacks, Defenses, and Future Directions](https://mlsanity.com/paper/2609.27847/) (2026)
- [RawVLA: Embodied Neural Image Signal Processor For Robotic Manipulation](https://mlsanity.com/paper/2609.37530/) (2026)

## Abstract (arXiv)

Current Vision-Language-Action (VLA) models rely primarily on RGB perception, preventing them from capturing modalities such as thermal signals that are imperceptible to conventional visual sensors. Moreover, end-to-end generative policies lack explicit safety constraints, making them fragile when encountering obstacles and novel scenarios outside the training distribution. To address these limitations, we propose Safe-Night VLA, a multimodal manipulation framework that enables robots to see the unseen while enforcing rigorous safety constraints for thermal-aware manipulation in unstructured environments. Specifically, Safe-Night VLA integrates long-wave infrared thermal perception into a pre-trained vision-language backbone, enabling semantic reasoning grounded in thermodynamic properties. To ensure safe execution under out-of-distribution conditions, we incorporate a safety filter via control barrier functions, which provide deterministic workspace constraint enforcement during policy execution. We validate our framework through real-world experiments on a Franka manipulator, introducing a novel evaluation paradigm featuring temperature-conditioned manipulation, subsurface target localization, and reflection disambiguation, while maintaining constrained execution at inference time. Results demonstrate that Safe-Night VLA outperforms RGB-only baselines and provide empirical evidence that foundation models can effectively leverage non-visible physical modalities for robust manipulation.