Mobile and Wearables

Context Awareness

Smartphones these days can do some pretty amazing things—from providing directions to the nearest cup of coffee to birthday and event reminders. But a context-aware phone is more than a map and calendar: it can become a personal valet. Through the many sensors that are available on today’s smartphones, we can personalize our devices to sense and learn our habits and preferences. Not only can your phone fetch you the location of the nearest cup of coffee, it can also tell you how long it will take to arrive. As more and more sensor data is collected and processed, the better phones will get at becoming context-aware of our needs and wants.

Context awareness can be expanded beyond our human behaviors to enable valuable connectivity with our environments—Internet of Things (IoT). An accelerometer used in an IoT application can learn the signature of how an individual opens a door, enabling it to “guess” who is entering the house. Your phone can know how often a pot of coffee has been lifted and for how long—and can then be configured to send a notification that it’s empty and ready to be refilled.

Factors that are essential to successful context awareness are the ability to accurately sense and the ability to continuously sense without prematurely draining the device battery. PNI’s SENtral sensor fusion coprocessor can asynchronously sample multiple sensors with accurate timing and batch the data while consuming almost no power. The algorithms embedded in SENtral are “tunable” to most effectively meet the needs of all applications—making higher levels of decision-making a reality.

Activity Monitoring

Activity-tracking applications and devices are commonplace – from devices worn on the body to biometric tracking built into the latest generation of mobile phones. Most of these are accelerometer-only based trackers, which are limited in what they track and are not consistently accurate. There is not much difference between accelerometers for consumer products. The real difference comes from the algorithms transforming the accelerometer readings into actionable data.

Fitness trackers, whether used by an endurance athlete or the casual user, are expected to log data for many hours—and still be able to provide useful information at the touch of a button when needed. Accelerometer-only data is the best for conserving power. PNI’s accelerometer-only step-counting algorithm running on SENtral uses a mere 20.6 µA  for a total system (SENtral + accelerometer) power draw of <26µA.

Adding gyroscope data to the sensor fusion algorithm allows for accurate distance traveled without user input (i.e., stride length) and reduces false and missed steps. The gyro helps accurately maintain both the instantaneous and long-term Earth frame reference for gravitational and linear accelerations.

Accelerometer + Gyroscope + Magnetometer:  Full 9-Axis Sensor Data

9-axis sensor fusion outputs available to a clever engineer can provide the data needed for gesture and complex motion tracking, such as a golf swing or swim stroke. Such advancements can provide beginner and advanced athletes with important information on their form to improve performance and prevent injury.

Power Consumption

The addition of more sensors can lead to a quickly draining battery. PNI’s SENtral sensor fusion coprocessor not only provides the data needed for advanced motion tracking, it can also do so without having to wake up a system MCU. Algorithms are available to tap into the different sensors only when needed, providing the required power savings.

Related Products