Infrared Hotspot-Guided Early Warning of Lithium-Ion Battery Thermal Runaway Under Mechanical Abuse

Electrical Engineering and Systems Science > Systems and Control
[Submitted on 29 Jun 2026]
Title:Infrared Hotspot-Guided Early Warning of Lithium-Ion Battery Thermal Runaway Under Mechanical Abuse
View PDF HTML (experimental)Abstract:Mechanical abuse can trigger thermal runaway (TR) in lithium-ion batteries through localized heat generation before sensor signals become decisive. This paper proposes a two-stage early-warning approach that estimates localized thermal instability from infrared hotspot dynamics and then fuses this instability score with mechanical, electrical, thermal, and image-intensity features for a 20-frame warning horizon. Evaluation uses repeated experiment-wise three-fold validation, with out-of-fold Stage-I scores during Stage-II training to prevent stacked-model optimism. Hotspot dynamics alone achieve Stage-I ROC-AUC 0.945, and the two-stage classifier reaches Stage-II ROC-AUC 0.908, exceeding direct multimodal fusion while preserving an interpretable intermediate instability signal. Thermal gradient rise precedes voltage-based detection by 40 frames (4 seconds) on average, enabling earlier battery management system intervention. Lead-time analysis at a fixed 0.5 threshold yields a 14.8-frame mean lead time.
Current browse context:
eess.SY
References & Citations
Loading...
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Connected Papers (What is Connected Papers?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
alphaXiv (What is alphaXiv?)
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Hugging Face (What is Huggingface?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

* Mechanical abuse can trigger thermal runaway in lithium-ion batteries through localized heat generation.
* A two-stage early-warning approach is proposed based on infrared hotspot dynamics to estimate thermal instability.
* This instability score is fused with mechanical, electrical, thermal, and image-intensity features for a 20-frame warning horizon.
* The evaluation used repeated experiment-wise three-fold validation with out-of-fold Stage-I scores during Stage-II training.
* Hotspot dynamics alone achieved a Stage-I ROC-AUC of 0.945.
* The two-stage classifier reached a Stage-II ROC-AUC of 0.908.
* Thermal gradient rise precedes voltage-based detection by an average of 40 frames (4 seconds).
* Lead-time analysis at a fixed 0.5 threshold yielded a mean lead time of 14.8 frames.

A two-stage early-warning approach is proposed to detect thermal runaway (TR) in lithium-ion batteries caused by mechanical abuse, leveraging infrared hotspot dynamics. The initial stage estimates localized thermal instability from the hotspot data, which is then fused with mechanical, electrical, thermal, and image-intensity features for a 20-frame warning horizon. Evaluation involved repeated experiment-wise three-fold validation, incorporating out-of-fold Stage-I scores during Stage-II training to mitigate stacked-model optimism. Hotspot dynamics alone achieved a Stage-I ROC-AUC of 0.945, and the complete two-stage classifier reached a Stage-II ROC-AUC of 0.908, outperforming direct multimodal fusion while retaining an interpretable intermediate instability signal. The method demonstrated that thermal gradient rise precedes voltage-based detection by an average of 40 frames (4 seconds), suggesting potential for earlier battery management system intervention. Lead-time analysis at a 0.5 threshold yielded a mean lead time of 14.8 frames.

The research introduces a method for extracting predictive thermal warnings from complex, multimodal sensor data under conditions of mechanical stress, suggesting that the transient infrared hotspot dynamics offer an intrinsically valuable, intermediate signal beyond standard electrical or thermal measurements. The success in achieving a competitive classification score (0.908) via a staged approach highlights a potential advantage in separating features related to physical instability from correlated sensory inputs, which aligns with the goal of creating interpretable intermediate signals for control systems. The fact that hotspot dynamics precede traditional voltage-based detection by significant timeframes suggests a critical gap in current safety monitoring protocols concerning the temporal resolution and specificity of early thermal event detection. The pattern is one where novel fusion techniques successfully distill physical causality (hotspot) into actionable foresight, offering a means to operationalize predictive physics within battery management systems. The implication is that exploiting these fine-grained spatio-temporal dynamics during mechanical stress can significantly enhance system resilience by enabling pre-emptive intervention before catastrophic failure signatures manifest in conventional metrics.

Infrared Hotspot-Guided Early Warning of Lithium-Ion Battery Thermal Runaway Under Mechanical Abuse

Facts Only

Executive Summary

Full Take

Infrared Hotspot-Guided Early Warning of Lithium-Ion Battery Thermal Runaway Under Mechanical Abuse

Facts Only

Executive Summary

Full Take