MEMS vs. FOG: How should you choose an inertial navigation system?(Part 2)

Microelectromechanical systems (or MEMS) have grown rapidly since their early concepts in the 1950s. Made of tiny integrated circuits and silicon-based microelectronics, the technology has dramatically changed industrial and consumer electronics, including inertial navigation systems, creating a variety of inertial sensors including gyroscopes, accelerometers and magnetometers.

The main advantage of MEMS so far has been their extremely low price compared to their FOG counterparts, sometimes as much as 10 times lower for similar or lower performance. The use of cheaper materials, better processes, and smaller sizes all help make MEMS cheaper to produce and drive their adoption in applications where FOG costs are too high to make commercial sense, such as in-vehicle GPS, wireless Man-machine, or camera pointing.

MEMS devices are very small and light. While FOGs are relatively large and heavy and cannot be used in tight spaces such as smartphones and toys, MEMS are the perfect solution for space-constrained applications. MEMS are now ubiquitous, from consumer applications to industrial applications in a variety of industries. This small form factor is significantly driving the adoption of MEMS in the drone surveying market, especially lidar surveying, where higher accuracy is required while remaining relatively small and lightweight to fit drones is critical of.

MEMS also consume less power than FOGs, allowing longer mission times for fuel-constrained vehicles. Combined with their small size and light weight, MEMS is the solution of choice for many unmanned vehicles requiring the lowest possible SWaP-C (Size, Weight, Power and Cost).

MEMS accuracy is particularly prominent in predictable dynamic environments, where the overall behavior of the vehicle is predictable without sudden, drastic changes in attitude or orientation that would confuse the internal dynamic motion of the steering filter output data Model constraints. If assisted by external sensors (speed odometer, etc.), the overall solution can be well suited for ground vehicles, for example, which are relatively expected to change speed and direction.

MEMS are not without limitations though. Due to their mechanical properties and components that vibrate at high frequencies, MEMS are more sensitive to vibrations, especially at harmonic frequencies. Vibration adds noise to the sensor output signal, causing offsets that need to be corrected by software.

This question may have some practical consequences. A non-negligible number of drone gyroscopes have been found to have resonant frequencies in both the audible and ultrasonic frequency ranges, making them susceptible to loudspeaker noise. It is thus possible to crash a drone at a distance with a "sonic attack" using a speaker set to the correct frequency.

MEMS are also typically prone to g-sensitivity errors in gyroscope measurements due to linear acceleration, resulting in large biases that directly affect the accuracy of attitude estimates in INS. While in highly dynamic domains such as guided weapons and drones, accelerations are often short (just a few seconds) but intense (5g or more), the errors accumulated over time are not negligible and need to be compensated for. Correction is done at the filter level, but adds another level of complexity that FOG alternatives simply don't suffer from.

conclusion

Not all FOG and MEMS solutions are comparable to each other, as differences in price and performance may also be important. However, if you compare low-end FOG with high-end MEMS, there is still some comparability, and other factors such as size and application need to be considered.

Of the two systems, FOG will always provide the highest level of performance. The real question is: how much are you willing to pay for it?

Standard

Application