A long time after providing the Driving Do-Not-Disturb feature exclusively on the Pixel 2 family, Google is now expanding the use cases of its driving-detection technology by releasing a Transition API. The new API is aimed to enable third-party app developers to integrate driving-detection features into their creations. The Pixel-exclusive feature, which was spotted on the latest Pixel handsets in October last year, automatically turns on the Do Not Disturb mode when it detects users in a moving vehicle. The technology powering that feature essentially uses artificial intelligence (AI) to determine transition using device motion and Bluetooth connections.
Google says that the Transition API helps Android developers leverage the same sets of training data and algorithmic filtering that it used particularly for the Driving Do-Not-Disturb feature. Software company Intuit tested the functionality of the API on its QuickBooks Self-Employed app and found it significantly easier than an in-house solution that combines GPS, phone sensors, and other metadata to track car mileage for self-employed workers. "Before the Transition API, we created our own solution to track mileage that combined GPS, phone sensors, and other metadata, but due to the wide variability in Android devices, our algorithm wasn't 100 percent accurate and some users reported missing or incomplete trips," said Intuit engineers Pranay Airan and Mithun Mahadevan. "We were able to build a proof-of-concept using the Transition API in a matter of days and it has now replaced our existing solution, offering a more reliable solution that also reduced our battery consumption."
Similar to its adoption on the QuickBooks Self-Employed app, Life360, an app that offers a timeline of locations, trips, and drives, also adopted the Transition API to enhance driving recognition features. The API, among other benefits, helped to reduce battery consumption and enhance drive-tracking accuracy through the Life360 app.
The initial focus of the Transition API is limited in nature. However, Google is in plans to add context-aware features to enable distinction between road and rail vehicles in the coming months. In the meantime, developers can learn how they the new framework can be a part of their apps by visiting its dedicated developer guide. The guide mentions detection of activities such as walking, biking, and driving that all can be a part of many Android apps soon.