Remembering various combinations of usernames and passwords is a major challenge for most people, but despite this it is still the most prevalent way of identifying to online services.

Your cell phone knows you better than you know yourself. Various sensors capture your usage, when you are at home, when you are at work and when you are on vacation. The phone recognizes your fingerprint, voice and face. It knows what apps you use and how you use them. This and more is information that can be used to identify you in a secure manner. Yet, no one has so far put all this information together in a way that benefits you. The project aims to change this by leveraging the processing power of the smart device.

We will use all the knowledge in your phone, to make your life easier, without any information leaving your phone. You will no longer be required to remember your passwords on digital services, because your phone will know that it is you trying to log on, and take care of it.

The project will do this in a way that is both safe and firmly rooted in your identity, while your privacy is respected. Here, the mobile phone constitutes a unique opportunity, because the SIM card is linked to your identity, like many even today have done with their Bank ID. The phone is also something you practically always have with you.

The biggest challenge for this project is to find methods that allows the phone to sufficiently quickly detect that it is no longer in the your hands, and then just as quickly feel "safe" when it is back in the right hands.

The project’s approach will be to analyze different types of data from the user’s mobile phone, for example, sensor data, data about networks and position, and possibly app data. The software will compare current data from the phone to historical data, combined with analyzes based on machine learning, to decide if the mobile phone is under the control of the correct user at the moment. The project will carry out experiments with test users in several trials.

In the first six months, the project has developed a first version of an app for data collection of sensor data. In addition, sensor data have been analyzed using machine learning.