Integrating generative adversarial networks with IoT for adaptive AI-powered personalized elderly care in smart homes

Fawad Naseer; Abdullah Addas; Muhammad Tahir; Muhammad Nasir Khan; Noreen Sattar

doi:10.3389/frai.2025.1520592

Integrating generative adversarial networks with IoT for adaptive AI-powered personalized elderly care in smart homes

Front Artif Intell. 2025 Feb 13:8:1520592. doi: 10.3389/frai.2025.1520592. eCollection 2025.

Authors

Fawad Naseer¹, Abdullah Addas^{2

3}, Muhammad Tahir⁴, Muhammad Nasir Khan⁵, Noreen Sattar⁶

Affiliations

¹ Department of Computer Science and Software Engineering, Beaconhouse International College, Faisalabad, Pakistan.
² Department of Civil Engineering, College of Engineering, Prince Sattam Bin Abdulaziz University, Alkharj, Saudi Arabia.
³ Landscape Architecture Department, Faculty of Architecture and Planning, King Abdulaziz University, Jeddah, Saudi Arabia.
⁴ Department of Computer Software Engineering, Sir Syed University of Engineering and Technology, Karachi, Pakistan.
⁵ Department of Electrical Engineering, Government College University Lahore, Lahore, Pakistan.
⁶ Computer Science Department, University of Agriculture Faisalabad (UAF), Faisalabad, Pakistan.

Abstract

The need for effective and personalized in-home solutions will continue to rise with the world population of elderly individuals expected to surpass 1.6 billion by the year 2050. The study presents a system that merges Generative Adversarial Network (GAN) with IoT-enabled adaptive artificial intelligence (AI) framework for transforming personalized elderly care within the smart home environment. The reason for the application of GANs is to generate synthetic health data, which in turn addresses the scarcity of data, especially of some rare but critical conditions, and helps enhance the predictive accuracy of the system. Continuous data collection from IoT sensors, including wearable sensors (e.g., heart rate monitors, pulse oximeters) and environmental sensors (e.g., temperature, humidity, and gas detectors), enables the system to track vital indications of health, activities, and environment for early warnings and personalized suggestions through real-time analysis. The AI adapts to the unique pattern of healthy and behavioral habits in every individual's lifestyle, hence offering personalized prompts, reminders, and sends off emergency alert notifications to the caregiver or health provider, when required. We were showing significant improvements like 30% faster detection of risk conditions in a large-scale real-world test setup, and 25% faster response times compared with other solutions. GANs applied to the synthesis of data enable more robust and accurate predictive models, ensuring privacy with the generation of realistic yet anonymized health profiles. The system merges state-of-the-art AI with GAN technology in advancing elderly care in a proactive, dignified, secure environment that allows improved quality of life and greater independence for the aging individual. The work hence provides a novel framework for the utilization of GAN in personalized healthcare and points out that this will help reshape elderly care in IoT-enabled "smart" homes.

Keywords: IoT-enabled smart homes; adaptive artificial intelligence; generative adversarial networks (GANs); healthcare AI applications; personalized elderly care; predictive healthcare analytics; proactive health monitoring; synthetic health data.

Grants and funding

The author(s) declare that financial support was received for the research, authorship, and/or publication of this article. This work was supported by the Prince Sattam bin Abdulaziz University through the project number (PSAU/2024/01/99520).