Abstract
In high ambient temperatures, external elevated temperatures amplify the effect of internal power losses in an electronic device and can lead to device overheating. This disclosure describes techniques to estimate the ambient temperature of an electronic device using a machine learning (ML) model. The ML model is provided with inputs such as temperature sensor readings, battery drain rate, on-device power measurements, and device charging status, optionally including the charging method (wired or wireless). Using a representative training dataset, the machine learning model is trained to generate ambient temperature predictions in real-time. If high ambient temperature is detected, the internal power consumption of the device can be adjusted (e.g. through thermal throttling) to prevent the device from overheating.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Chandula, Sayanna; Babra, Jagga; Wang, Kame(TeYuan); and Balaji, S Ashwin, "Machine Learning Based Ambient Temperature Detection to Prevent Device Overheating", Technical Disclosure Commons, (October 22, 2025)
https://www.tdcommons.org/dpubs_series/8759