Context-aware: How PCs are leading the way

Adapting professional PCs to the hybrid work model

Imagine a world where sensors could do more than just collect data. At the time of its release, ST’s motion sensor was a revolutionary device, being the first to integrate AI. Highly compact, this ‘always‑on’ sensor, also called LSM6DSOX, observes our environment, process data, and runs machine learning algorithms. To achieve this high level of autonomy and intelligence, the motion sensor relies on a tiny but powerful hardware block: a machine learning core.

The best part? The block offloads algorithms from the application processor to the motion sensor itself.

And this migration certainly changed the game for the personal electronics industry. By enabling the implementation of context aware features while maintaining energy efficiency, edge AI laid the foundation for a new generation of smart devices.

Edge AI: a key enabler for context-aware features

HP worked closely with ST’s experts to develop and train AI models that recognize different user activities based on device and user motion. Several use cases were addressed, including scenarios where the laptop is placed on a table, on the user’s lap, or carried inside a bag and taken out. Various groups within HP volunteered to conduct these activities to analyze and collect these unique motion patterns, which then were fed into a machine learning tool. This helped create an AI model specific to HP devices. The machine learning core—embedded in the sensor— executes this model, allowing HP computers to detect changes in the user’s scenario and respond accordingly.

ST's AI-enabled motion sensor with machine learning core

Now imagine a sensor with edge AI running at 34 microamps. This is precisely what ST’s motion sensor has brought to developers. The LSM6DSOX was in 2020 a brand‑new AI‑enabled sensor packed with an inertial measurement unit featuring a high‑performance accelerometer and gyroscope.

With this sensor onboard, HP developers were able to create a new line of context‑aware laptops—EliteBook 1000— with minimal impact on battery life.

The simplicity of the machine learning core is a key element in achieving an unparalleled level of low power. The hardware block allows for the running of up to eight decision trees in just five simple steps: capture data, label it, build a decision tree with free tools, embed it in your application and finally process data in real time. With this streamlined method—yet ultra reliable—each new tree requires a power budget equivalent to that of a sleep mode. This enables the application processor to stand by and save power for what really counts.

Context‑awareness has already proved its worth. Carving out its status as a global leader in PCs, HP worked together with ST to set a new standard across the personal electronics industry.