A Wireless Micro Inertial Measurement Unit (IMU) - Embedded.com

A Wireless Micro Inertial Measurement Unit (IMU)

In recent years, location-based services for indoor mobile devices are becoming more and more popular due to their increasing functional range. In the context of indoor appli- cations, where GPS signals are not available, one promising localization approach, e.g., for cell phones, people, or mobile robots, relies on inertial sensors.

For human tracking, miniature inertial sensors integrated into clothes or shoes can extend the functionality of pedometers by full 3D location information. Even more complex tasks like full-body human motion capture have been addressed using inertial sensors. Additionally, the combination of inertial sensors with wireless signal sensors allows for correction of the errors accumulated during integration of inertial sensor data as shown by Xsens with the MVN MotionGrid.

Similarly, personal devices like cell phones can be localized inside of buildings. In the domain of mobile robots inertial sensors are commonly used in com- bination with cameras for fiight stabilization or autonomous hovering of helicopters or quadrotors. Furthermore, inertial sensors have been applied for localization of airplanes and miniature indoor blimps.

For application in such devices or on robots, small commercially available inertial measurement units (IMUs) typically incorporate MEMS acceleration sensors, gyroscopes, and magnetometers measuring in three axis. Additionally, most of the commercially available IMUs offer a measurement data preprocessing and robust on-board sensor data fusion for 3D orientation. However, due to their size, weight, or power consumption, these IMUs are not ideally suitable for certain applications such as miniature aerial vehicles or inconspicuous integration into clothes or shoes.

To date the commercially available IMUs are not implementable in small mobile objects due to the constraints like size, high power consumption and portability issues. In contrast to the commercial products, existing research prototypes are smaller and lighter but the data performance is a bottleneck. For the data processing it is important to have high raw data rates.

In this paper, we present a wireless micro inertial measurement unit (IMU) with the smallest volume and weight requirements available at the moment. With a size of 18 mm x 16 mm x 4 mm, this IMU provides full control over the data of a three-axis accelerometer, a three-axis gyroscope, and a three- axis magnetometer. It meets the design prerequisites of a space- saving design and eliminates the need of a hard-wired data communication while still being comparable to state of the art commercially available MEMS IMUs.

For reading the digital sensor data and radio data trans- mission a CC430 microcontroller is used, which combines an integrated CC1101 wireless transceiver and an MSP430 microcontroller. The MSP430 is a 16-bit RISC mixed-signal processor for ultra low power applications.

The CC430 microcontroller sends the collected raw data to a base station wirelessly with a maximum sensor sample rate of 640 samples per second. Thereby, the IMU performance is optimized by moving data post processing to the base station. This development offers important features in embedded microsystem applications with their significant size and weight requirements.

In our methodology of designing the Micro-IMU, we combined the aim of an applicable IMU for embedded microsystems and improved the characteristics in size, weight, and power consumption while the performance is still comparable to state-of-the-art commercially available MEMS IMUs.

By using MEMS sensors with large scale integration this becomes possible. Modern accelerometers and gyroscopes feature three-axis technology and integrated analog-to-digital conversion with automatic temperature compensation in a one- chip-design. This saves space in the IMU design as analog converters, high precision voltage references and 3D packaging are not necessary. Thus, a four-layer PCB is sufficient for integration of all sensors and a microcontroller.

To read more of this external content, download the complete paper from the online open archives at the University of Frieburg.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.