Movea's sensor funsion technology is found inside a range of embedded devices including Babolat's Play and Connect Tennis racquet. It is also found in Freescale's multisensor USB development platform (see Sensor Fusion Or Sensor Confusion?). It uses Freescale's ColdFire microcontroller and MEMS sensors along with Movea's SmartMotion software to provide integrated sensor information using Microsoft's Windows 8 new USB sensor queue interface.
Movea's MotionCore software (Fig. 1) runs on most low power microcontrollers to provide sensor fusion supporting a range of sensor such as accelerometers, gyroscopes, and magnetometers. MotionApps can utilize Movea's API. MotionCore can also handle standard interfaces like the Windows 8 support.
Figure 1. Movea's MotionCore provides integrated sensor information.
Movea has a client/server architecure. The SmartMotion firmware ties multiple sensors together and communicates with a SmartMotion server. This connection can be via a range of interfaces from RF4CE to WiFi and Bluetooth to USB. The SmartMotion Services API provides the interface to the server.
Movea's SmartMotion technology has been used in a range of products including MoveTV. MoveTV provides a gesture-based interface using a sensor enhanced TV remote control. Movea provides a Gesture Builder application so developers can create a database of their own gestures.
Sensor fusion provides a range of advantages. Lower power sensors like accelerometers can be used to simulate other on-board sensors that can be powered down when lower accuracy results are needed. Likewise, a microcontroller-based solution can provide improved power management and sensor control. For example, some sensors like the gyroscope may be turned off if the device is not moving since there is no change of information. An accelerometer can detect movement that would then activate other sensors.
The microcontroller can also improve overall system operation. One example is support for dead reckoning that is often needed indoors were there are no additional references such as GPS. A single sensor is often inadequate for good dead reckoning support past a couple feet. Combining sensor data can significantly improve these results. An experiment of over a distance of 80m resulted in an average error under 3% and a step count error under 1%. That is actually quite good because normally cummulative error will push the error rate to 100% after a short period of time.
This embedded architecture is available to developers and it has also found a home in a number of applications areas. Movea's SmartMotion for sports and eHealth is one of these areas. Typical sensor configurations are 3A3M and 3A3G3M (3D acceleromenter, 3D gyroscope, 3D magnetometer). These combinations are used in medical rehabilitation applications.
SmartMotion technology has been used in Babolat's Play and Connect tennis racquet (Fig. 2) that combines MEMS sensor, wireless connectivity and Movea's sensor fusion software for real-time tracking. It provides feedback on the players swing.
Figure 2. Babolat's Play and Connect tennis racquet combines MEMS sensor, wireless connectivity and Movea's sensor fusion software for real-time tracking.
Babolat's Play and Connect tennis racquet delivered it sensor information to tablets (Fig. 3) via wireless connection. The technology was first shown at the Roland-Garros French Open competition in May 2012.
Figure 3. Babolat's Play and Connect tennis racquet can deliver the MEMS sensor data to provide real-time player analysis.
The racquet has a battery, microcontroller and sensor complex in the handle. The micro analyzes sensor information and transmits it via a Bluetooth transceiver. A courtside receiver can be linked to a more powerful computer and network interface. It is expected to add about $100 to the cost of a racquet.
