You are wandering around a shopping mall, looking for a specific store. You can find the store’s website on your phone, but unfortunately the phone is not yet able to guide you to its door. You locate a display board in the mall to show you the way. Outside the mall, your phone could guide you right to the door of the store, but the satellite navigation (GNSS) signal is not available inside, as it cannot penetrate the ceiling and the walls.

When you finally return to the parking hall and manage to find your car without phone navigation, you get in the car and enter your next destination into the car navigator. However, today’s car navigators cannot show you the way out of the parking hall and onto the road, and they are often unable to tell you which way to turn when exiting the hall, as even though they know the location of the building, they cannot tell the direction in which the car is moving. Luckily, today’s human drivers are able to read signs and usually determine the right way, even in unfamiliar places.

In a few years, an autonomous car will be a bit more lost in a similar situation and may occasionally turn to the wrong direction before finding the GNSS satellites and consequently the right direction. It could be quite awkward to sit in a robot taxi and watch an autonomous car take a wrong turn during rush hour while knowing yourself that it will soon have to make an extra U turn or circle the block.

An ordinary compass is not enough

In the future, problems such as the ones described above will hopefully be a thing of the past, thanks to Murata’s development pioneering. At the moment, Murata’s R&D unit is carrying out development work to facilitate and develop a new kind of MEMS gyrocompass. A MEMS gyrocompass is a compass based on movement sensors that utilizes MEMS angular velocity sensors (gyros) to determine the direction of Earth’s axis, i.e., true north.

To date, more conventional gyrocompasses have only been available for applications such as aviation and seafaring, in which their large size and high price have been justified. The new MEMS gyrocompass is intended to reduce both the size and the price by several orders of magnitude, which is believed to be enough in the initial phase for the autonomous vehicle and industrial application markets at least. The markets for consumer electronics, such as smartphones, require a further reduction in terms of size and price alike.

What about conventional compasses? They also show the direction, right? Affordable electronic magnetic compasses have been on the market for more than a decade – in devices such as smartphones. A compass based on measuring Earth’s magnetic field works well when orienteering in the woods, but in most everyday applications, its use faces fundamental problems, such as magnetic variation and magnetic fields in the environment.

Earth’s magnetic poles deviate quite a bit from the true north, and the difference is referred to as magnetic declination. The declination depends on the location, in addition to which it changes over time as the magnetic poles constantly move at an increasing speed. However, environmental disturbances are an even bigger problem. The environment modified by humanity interferes with electronic compasses through a variety of magnetic fields and magnetic materials.

Several materials in the environment, such as car chassis, reinforced concrete and other metallic structures, are magnetic, while power lines, the large electric currents of car electronics or battery chargers, for example, generate magnetic fields that can easily be several times greater than Earth’s small magnetic field, even at a distance. In light of the foregoing, a magnetic compass is particularly unsuited for determining the direction of a vehicle.

Making determining the direction possible

A MEMS gyrocompass usually operates on the principle of finding north as the angular velocity sensor is turned in different directions in a plane. Turning the measurement axis of the sensor in a plane substantially eliminates the zero error of the angular velocity sensor, which in MEMS angular velocity sensors is typically several hundred times greater than the angular velocity depending on the direction itself. The reading of the sensor should change sinusoidally as a function of direction; when the measurement axis points north, the sensor sees a portion of Earth’s angular velocity as an increase in the reading and, correspondingly, as a decrease when pointing south.





Finding north in a plane. 

At its best, the current performance of Murata’s MEMS gyro technology makes it possible to find the direction of north with a ±4° accuracy in less than 10 seconds. The time required for determining the direction increases to the square of the accuracy, whereby a ±1° accuracy takes 16 times the time, i.e., slightly over 2 minutes. Due to this strong time dependency, the application in question has only recently become possible in practice as the noise of the best affordable MEMS gyroscopes has reached a critical level.

Murata is a market leader in the performance of the movement sensors of vehicle stability control systems. The noise level of a typical angular velocity sensor on the market is 4–10 times higher than that of Murata’s new third-generation sensors. This leads to measurement times that are tens of times longer, taking several minutes instead of seconds. From a safety standpoint alone, such long measurement times are not feasible in traffic. The significant performance advantage gives us a unique opportunity to expand into new application areas, which is not yet even possible in practice with competing MEMS sensors.

In the future, when the prices and sizes of high-performance angular velocity sensors are reduced enough, MEMS gyrocompasses will find their place in our personal devices as well. Then, thanks to this technology, you will be able to find the store you were looking for easily with your phone and without having to look around.

Demonstration of Murata’s MEMS motion sensor based gyrocompass introduced at the CES 2018 fair. 


Anssi Blomqvist,
Fellow, R&D