Smartphones Use Night Hours for Background Processes

Smartphones perform a variety of background processes during nighttime charging that are unknown to most users. These activities begin as soon as the device is connected to power and Wi-Fi while in sleep mode. A central component of these processes is the internal service called JobScheduler, which executes tasks specifically while the device is not actively in use. An example of these background activities is the AI-driven enhancement of the keyboard app Gboard. During charging, a small AI model is downloaded from the internet and trained locally on the device.

This model analyzes the user's typing habits to optimize word prediction. No actual texts are transmitted; instead, mathematical gradients representing abstract learning outcomes are sent. The method used here is called Federated Learning. It aims to protect user privacy by processing only anonymized data. However, researchers have found that under certain circumstances, it is possible to reconstruct specific words from these gradients.

Google states that the transmitted data is additionally encrypted and mixed with data from other users. Another process that occurs during charging is the analysis of photos in the gallery app. When the smartphone is charging, the media scan service starts, capturing all new, modified, or deleted files. A local AI model analyzes the photos for facial features and creates mathematical profiles to group images of the same person. Additionally, photos are tagged based on their content, making them easier to find through text search.

Screenshots are also recognized and made searchable through optical character recognition (OCR). Photos that are blurry or poorly lit are marked as candidates for deletion, contributing to storage optimization. The continuous use of smartphones leads to an accumulation of invisible data over time. To counteract this, the system defragments databases and clears storage while the device is charging in sleep mode. These processes are designed to optimize device performance and enhance the user experience.

For users who want to limit their nighttime data transfers, there are various tips and settings explained in an accompanying article. These measures can help reduce the amount of data sent in the background and protect user privacy. The use of background processes during charging is an example of the advanced features of modern smartphones. These technologies are designed to improve the user experience and make devices more efficient. According to Google, these features are continuously being developed to meet user needs. The use of Federated Learning and other AI technologies demonstrates how smartphones can adapt and learn to optimize interaction with users. Google has announced that these technologies will also be further refined in future updates.