GNSS, dead reckoning, and MEMS IMUs

GNSS is a wonderful invention, and low cost receivers have crept into smartphones and other mobile devices. However, GNSS does not solve all problems, especially in urban environments. The canyon effect blocks signals at street level between tall buildings, and signals do not penetrate to the interior of parking garages, tunnels, basements or lower floors of large structures.

When aircraft or ships want to know their exact position, they rely on sophisticated inertial measurement units (IMUs), combining accelerometers and gyroscopes for all three axes. State-of-the-art IMUs often use a ring laser gyroscope, with almost no moving parts compared to mechanical spinning gyros. This results in near-zero drift, and the ability to operate in any orientation. Ring laser gyros rely on two counter-rotating beams of light and the Sagnac effect, tracking small shifts in standing wave interference patterns.

A computer performing dead reckoning calculations usually backs IMUs. Dead reckoning is the process of estimating current location based on a prior location fix and any movements executed since. For instance, GPS can provide an initial fix on a map location, while dead reckoning algorithms continue to compute location from accelerometer and gyroscope readings.

GNSS also experiences latency, sometimes a little due to signal travel back and forth to satellites, and sometimes a lot due to computations needed to get a fix. In order to provide high accuracy and low latency, most implementations combine an IMU and GNSS with Kalman filtering or another advanced algorithm. GNSS augmentation also allows zeroing out accumulated IMU errors from many sources, preventing Kalman filter divergence.

Don’t have the space or the pocket change for a ring laser gyro or a mil-spec computer? Fortunately, the combination of MEMS sensors, low cost GNSS receivers, and MCUs have advanced enough to provide IMUs and dead reckoning computations able to keep position accurately enough for many applications.

This combination puts ST Micro solidly in the IMU business, with applications from cars to drones to robots. MCU lines based on ARM Cortex-M, including the latest STM32 F7 with the Cortex-M7 core, are well known. ST is also in the GNSS business. Earlier this year, they announced the Teseo III single chip receiver, with support for positioning networks including GPS (US), GLONASS (Russia), Galileo (Europe), and BeiDou (China).
The race has been on between MEMS sensor manufacturers to produce a smaller IMU, with six-axis integration. ST has just taken the lead with introduction of a 3mm x 3mm LGA automotive-grade sensor, the ASM330LXH. It isn’t their first part that small – the LSM6DB0 features a six-axis IMU with a Cortex-M0 controlller. However, the ASM330LXH is AEC-Q100 qualified with a temperature range of -40 to +85C, connecting to an MCU via SPI.

According to IHS, ST leads in market share for both consumer devices and automotive MEMS as of 1H 2014. With over 5 billion MEMS sensors shipped, ST is well poised for future growth. For more on MEMS technology and companies, visit the MEMS Industry Group.