Experimental Analysis of Thermal Behavior and Computational Workload in AI-Enabled Wearable Devices

  • Authors

    • Jenifar Devasagayam Independent Researcher, USA. Author

    DOI:

    https://doi.org/10.67228/30716357/IJMRSE-V9I2P102

    Published 04-09-2026

  • Artificial Intelligence, Wearable Devices, Thermal Behavior, Computational Workload, Edge Computing, Energy Efficiency, Inference Latency, Embedded Machine Learning, Temperature Monitoring, Thermal-Aware Scheduling

    Issue

    Vol. 9 No. 2 (2026)

    Section

    Articles

  • Abstract

    AI-powered wearables are increasingly performing increasingly complex functions such as physiological signal classification, activity recognition, anomaly detection and personalised health monitoring. But high workloads of inferences may also raise the temperature, energy consumption, and time for processors to complete their tasks, which can impact user comfort, battery life, and system reliability. The authors introduce a framework for experimental analysis of the correlation between the computational load and thermal performance of a wearable device powered by AI. The proposed methodology combines the workload characterization, temperature monitoring, utilization measurement of processors, estimation of power consumption, and inference performance evaluation under various operating conditions. Representative AI workloads are classified based on their computational complexity, model execution frequency and memory requirements. The performance measurement of thermal response includes temperature rise, heating rate, steady state temperature, and thermal recovery time; computational performance includes inference latency, throughput and energy efficiency. A controlled experimental protocol and a well-documented analytical framework are created to evaluate lightweight and intensive models. No physical measurement data are provided, so the results obtained are not reported as experimental findings. Rather, the article provides a strict foundation for future validation of hardware and thermal-aware AI workload management within wearable computing systems.

  • References

    [1] V. Sze, Y.-H. Chen, T.-J. Yang, and J. S. Emer, “Efficient Processing of Deep Neural Networks: A Tutorial and Survey,” Proceedings of the IEEE, vol. 105, no. 12, pp. 2295–2329, Dec. 2017, doi: 10.1109/JPROC.2017.2761740.

    [2] “Thermal Acceptability Limits for Wearable Electronic Devices,” in Proceedings of the 37th Semiconductor Thermal Measurement, Modeling & Management Symposium (SEMI-THERM), 2021.

    [3] A. Zavareh et al., “Soft Wearable Thermal Devices Integrated with Machine Learning,” Advanced Materials Technologies, 2023, doi: 10.1002/admt.202300206.

    [4] R. Kumar and B. Ghoshal, “Machine Learning Guided Thermal Management of Open Computing Language Applications on CPU-GPU Based Embedded Platforms,” IET Computers & Digital Techniques, vol. 17, no. 1, pp. 20–28, 2023, doi: 10.1049/cdt2.12050.

    [5] Brahmandam, L. M. K. (2024). An Empirical Evaluation of the Medallion Architecture on Databricks and Apache Spark with Snowflake: Throughput, Latency, and Cost for Batch and Real-Time Ingestion Patterns. International Journal of AI, BigData, Computational and Management Studies, 5(3), 197-206. https://doi.org/10.63282/3050-9416.IJAIBDCMS-V5I3P122

    [6] M. R. Stan, K. Skadron, M. Barcella, W. Huang, K. Sankaranarayanan, and S. Velusamy, “HotSpot: A Dynamic Compact Thermal Model at the Processor-Architecture Level,” Microelectronics Journal, vol. 34, pp. 1153–1165, 2003, doi: 10.1016/S0026-2692(03)00206-4.

    [7] K. Skadron, M. R. Stan, K. Sankaranarayanan, W. Huang, S. Velusamy, and D. Tarjan, “Temperature-Aware Microarchitecture,” ACM Transactions on Architecture and Code Optimization, vol. 1, no. 1, pp. 94–125, 2004, doi: 10.1145/980152.980157.

    [8] N. D. Lane, S. Bhattacharya, P. Georgiev, C. Forlivesi, L. Kawsar, and F. Kawsar, “DeepX: A Software Accelerator for Low-Power Deep Learning Inference on Mobile Devices,” in Proceedings of the 15th ACM/IEEE International Conference on Information Processing in Sensor Networks (IPSN), 2016.

    [9] S. Han, H. Mao, and W. J. Dally, “Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding,” in Proceedings of the International Conference on Learning Representations (ICLR), 2016.

    [10] Arora, A. S., Yachamaneni, T., & Kotadiya, U. (2021). Optimizing Multi-Tenant Resource Allocation in Cloud-Based Distributed Systems for Large-Scale AI Model Training Using In-Memory Computing. International Journal of Emerging Research in Engineering and Technology, 2(1), 37-46.

    [11] A. G. Howard et al., “MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications,” arXiv:1704.04861, 2017.

    [12] B. Jacob et al., “Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2018, pp. 2704–2713.

    [13] P. Warden and D. Situnayake, TinyML: Machine Learning with TensorFlow Lite on Arduino and Ultra-Low-Power Microcontrollers. Sebastopol, CA, USA: O’Reilly Media, 2019.

    [14] Paruchuri, J. K. (2022). Survey of cloud-native workflow orchestration with Apache Airflow. International Journal of Computer Techniques (IJCT), 9(3), 1-15.

    [15] C. Banbury et al., “MLPerf Tiny Benchmark,” in Proceedings of the NeurIPS Datasets and Benchmarks Track, 2021.

    [16] R. P. Dick, “Temperature-Aware Design and Management of Embedded Systems,” in Proceedings of the International Symposium on Low Power Electronics and Design (ISLPED). This topic provides relevant background on thermal constraints and power-aware embedded-system design.

    [17] Konatham, M. R., Uddandarao, D. P., Vadlamani, R. K., & Nair, R. (2025, September). Optimizing Credit Risk Assessment in Digital Banking Through AI-Driven Predictive Science. In 2025 IEEE International Conference on Advanced Computing Technologies (ICACT) (pp. 478-483). IEEE.

    [18] Gantikota, S. (2024). Mitigating OWASP Top Ten Risks in Cloud-Native Healthcare and Education Platforms: A Comparative Analysis of SQL Injection and Cross-Site Scripting Defenses. American International Journal of Computer Science and Technology, 6(1), 65-70. https://doi.org/10.63282/3117-5481/AIJCST-V6I1P107

    [19] “A Review on TinyML: State-of-the-Art and Prospects,” Journal of King Saud University – Computer and Information Sciences, 2022.

    [20] “Green Edge AI: A Contemporary Survey,” Proceedings of the IEEE, vol. 112, no. 7, pp. 880–911, 2024, doi: 10.1109/JPROC.2024.3437365.

    [21] Sandra, K. (2024). THE REGULATED BANKING AI LAKEHOUSE. INDO-CONTINENTAL ACADEMIC PUBLISHERS.

    [22] Sunkara, R. (2024). Hardware-in-the-Loop Power Profiling Automation for Consumer Streaming Devices: A Multi-Lab Framework for Regulatory Compliance Validation. International Journal of Emerging Trends in Computer Science and Information Technology, 5(4), 187-191. https://doi.org/10.63282/3050-9246.IJETCSIT-V5I4P121

  • Downloads