An Intelligent Tracking Model for Stolen Android Devices with Multi-User Authentication and Predictive Recovery Support

Abstract

Smartphone theft and loss continue to present major cybersecurity, privacy, and financial risks due to the increasing dependence on mobile devices for communication, digital transactions, identity management, and storage of sensitive information. Although existing recovery solutions such as Google Find My Device and similar commercial platforms provide basic device recovery functionalities, their effectiveness remains limited under real-world theft conditions due to reliance on continuous internet connectivity, enabled GPS services, and single-user authentication mechanisms. This study proposed and evaluated an intelligent Android device recovery model integrating multi-user authentication, offline tracking resilience, stealth-based monitoring, and lightweight machine learning-based trajectory prediction. The study adopted a Design Science Research methodology involving problem identification, framework design, prototype implementation, and performance evaluation. The proposed framework was implemented within an Android environment using Kotlin, Firebase Firestore, Google Maps API, and FIDO2/WebAuthn passkey authentication technologies. A K-Nearest Neighbors predictive subsystem was integrated to analyze historical movement behavior and estimate probable future device trajectories during recovery operations. The system further incorporated offline location caching, delayed cloud synchronization, trusted contact recovery support, and heatmap-based movement visualization. Evaluation findings demonstrated improved operational resilience, authentication reliability, predictive recovery support, and tracking continuity compared to conventional smartphone recovery systems. The integrated predictive subsystem achieved acceptable trajectory estimation accuracy with low Root Mean Square Error (RMSE) values, while offline synchronization mechanisms improved recovery continuity during unstable network conditions. The findings established that collaborative multi-user recovery strengthened recovery and operational flexibility. The study concludes that integrating predictive analytics, collaborative authentication, and offline operational resilience significantly enhances smartphone recovery effectiveness under real-world-oriented conditions and contributes meaningfully to intelligent mobile security research.