System and Method for Continuous Wearable User Authentication
Tech ID: 18-013
Inventors: Christian Poellabauer, Sudip Vhaduri
Date Added: December 10, 2020
Overview
A reliable authentication method using physiological data collected on smart wearable devices.
Technology Summary

The capabilities of wearable technologies continue to grow and presently include tracking and collecting physiological data, accessing and controlling other devices, accessing digital payment methods, and user identification. However, with increased capabilities, greater data storage capacity, and expanding interactivity, security challenges have also grown. First order security lies in limiting physical access to wearable technologies to authorized users. Current methods for doing so on mobile devices largely rely on one-time authentication methods such as passwords, facial recognition, or fingerprints.
Researchers at the University of Notre Dame have developed a continuous and reliable authentication method for wearables using data from physiological sensors and activity monitors. This implicit authentication does not require direct user input and relies on data easily obtained in these devices. Utilizing machine learning modeling and statistical analysis, the physiological data collected by the device is enough to continuously provide accurate identification of the user. Upon detecting unauthorized use, the device can deactivate services, prevent accessibility to private data, and notify the user. For shared devices, it can detect individual users and personalize features and services based on the preferences of the different users.
Market Advantages
- Of analysis from 421 FitBit users, 93% identification accuracy (sedentary) and 90% identification accuracy (non-sedentary)
- Convenient- no user input is needed
- Low-cost and low-energy
Applications
- Health insurance companies
- Life insurance companies
- Healthcare
Technology Readiness Level
TRL 3 – Experimental Proof of Concept
Publications
Wearable device user authentication using physiological and behavioral metrics. doi:10.1109/PIMRC.2017.8292272
Multi-Modal Biometric-Based Implicit Authentication of Wearable Device Users. doi:10.1109/TIFS.2019.2911170
Contact
Richard Cox
rcox4@nd.edu
574.631.5158