Kit Simplifies Sensor Hub Evaluation

March 25, 2014

Movea worked with STmicroelectronics to deliver a SASH (smart architecture sensor hub) evaluation kit with Movea's SmartMotion technology that can provide improved pedestrian navigation.

William G. Wong

Integrating sensor data into useful information is not just adding a bunch of numbers together or running a basic transform or correlation on some data. A sensor hub needs to handle a wide range of sensor information and be able to present that basic information as well as extrapolated information to an application that does not have to worry about how that information is obtained. That is what Movea's latest SASH (smart architecture sensor hub) evaluation kit does. It targets OEMs that will incorporate the context awareness technology (Fig. 1) into their products.

Electronicdesign Com Sites Electronicdesign com Files Uploads 2014 03 106231 Fig1sm

Figure 1. Movea's sensor hub SmartMotion technology can provide context-awareness to applications while simplifying the type and amount of data provided.

Electronicdesign Com Sites Electronicdesign com Files Uploads 2014 03 106231 Fig2sm

Figure 2. The sensor hub module fits inside a smartphone and can be significantly reduced in size after prototyping and evaluation is complete.

The SASH-1 is based on STMicroelectronics’ STM32F401 Cortex-M4 microcontroller. It runs Movea’s SmartMotion technology. The kit is base on Android 4.4 that runs on a platform like a Nexus smartphone (Fig. 2). The standalone sensing system provides the 9-axis motion processing support required by Android. It does so with low-power sensors using under 100 µW and under 3 mW with high performance sensors.

Related Articles

Using less than 4 mW, the system can provide pedestrian level dead reckoning. The ability to provide pedestrian navigation without the need for GPS is very useful especially in conditions where GPS is unavailable such as underground or in many buildings. After 15 minutes of operation it still has a distance deviation under 5% and 2% for orientation. It can determine floor changes with a 99.7% accuracy. The system is designed for integration with WiFi and Bluetooth positioning via triangulation as well as map matching when a map is available.

The system also provides step cadency in addition to direction and distance information. The system can consume less than 1 mW by scaling down the processing and functionality. Still, it can provide information such as context awareness such as whether a person is in a car, train, plane or even an elevator. Of course, it can tell whether the person is walking, running or resting as well.

The sensor hub can live in the smartphone or in a wearable using wireless communication support such as Bluetooth (Fig. 3). This can reduce the communication overhead since the host is not handling the sensor integration. Low power operation is also critical for wearables where power scavenging may be used or where limited battery capacity is available.