We were inspired to create a new innovative way to make user authentication easier. As passwords and other security questions are often hard to remember and people often resort to writing them down on paper or even their phones without any security protections, we thought that using the concept of gesture recognition could be applicable to user authentication.
What it does
Upon login, users will have to option to "login with Motion" similar to "login to Google or Facebook". They will be redirected to their phones or watches and will be instructed to make a gesture while the device is in their hand. The device will detect the movement of the device following their unique and custom gesture or motion. If the real time motion data gathered from the device matches the originally, saved password, the user will be able to access their accounts!
How we built it
We built this with an Arduino 101 with an accelerometer and Arduino IDE. We obtained the 3D points and processed the authentication in this application. We also built a login-page using HTML/CSS hosted on Microsoft Azure.
Challenges we ran into
We ran into issues with processing the accelerometer data and developing parameters and the algorithm to determine if the password was accurate due to the noise and data acquisition rate of the Arduino. We also had problems interfacing our live-hosted website to the Arduino because we didn't realize till pretty late that we needed a wifi-enabled accessory to do this. Additionally, we had difficulties navigating and obtaining information from our database on Microsoft Azure.
Accomplishments that we're proud of
We created a hardware project finally! and learned a lot about the limitations of connecting our backend and project. We were able to determine a threshold to determine if the password waveforms were similar enough to the original to authentication.
What we learned
We learned the intricacies of Arduino as we debugged our project throughout the past 24 hrs. We also learned how to host our website live and create a database through Microsoft Azure. We also learned that cybersecurity is a hard problem to solve as there are so many components to consider throughout the process.
What's next for Motion
1) Link all components of project onto one platform. 2) Encrypt Motion password data to increase cybersecurity. 3) Create a more comprehensive filter for the Arduino data.