Unmanned Systems Technology 036
38 and complexity can be reduced by manufacturing sensing elements and electrodes together as monolithic blocks. That can also increase capacitive sensitivities as well as linearity of output from capacitance to acceleration and angular rate data. Automating advanced component bonding methods such as seam-welding the MEMS devices to their boards can also increase resistance to harsh environmental conditions as well as product lifespans. Fibre optic gyros While MEMS devices still hold the advantage in size, weight and power, FOGs continue to have a modicum of superiority in terms of performance. Currently, the most cost-effective high- performance FOG IMUs have bias instability levels no higher than 0.1 º /h in all conditions – a level of accuracy that is now just about achievable by the best commercially available MEMS IMUs. However, examining the bias repeatability is important for understanding the continued difference in performance between FOG and MEMS IMUs. Bias repeatability refers to the average change in bias that occurs between each instance of the inertial sensor being switched on, and is a critical gauge for minimising long-term drift (and ensuring long-term reliability) of gyroscopic measurements. Among the various FOG designs available for unmanned systems, bias stability and bias repeatability tend to run numerically close together. For MEMS IMUs, however, bias repeatability tends to fall short of bias stability, varying across tens and sometimes hundreds of degrees per hour. To minimise bias repeatability requires very stable devices as well as good control over their mechanical, thermal and electrical qualities – more so than is possible within the physics of commercially available MEMS devices. Therefore, instead of never switching off their MEMS devices, which would align long-term drift more towards bias stability than bias repeatability, users of MEMS IMUs are recommended to carry out recalibrations ‘on the fly’ after a given number of hours of use. FOG users, by contrast, can omit this step from their maintenance checklists, as most FOGs for autonomous vehicles have bias repeatabilities of 0.1 º /h. This difference between MEMS IMUs and FOGs comes down largely to the laws of physics and the difference in size between their sensing structures and techniques. A bigger system yields larger and therefore more detectable signals that can be translated into inertial measurements – precisely the same quality that ties the accuracy and reliability of different FOGs to the length and amount of optical fibre used in their sensing coils. Specifically, the Sagnac effect (which is key to the interferometry by which FOGs measure changes in angular rate) is amplified by increasing the amount of fibre inside the FOG enclosure. With a greater volume of fibre for emitted light to travel through, the two emitted beams will undergo a greater delay differential in their travelled paths during rotations in their shared axis. This increase in interference between them means a bigger effect – and, by extension, a bigger range of effects – for the FOG processor to draw photometric measurements from. Of course, FOGs still take longer and cost more to manufacture, as they have a longer list of parts than MEMS devices and are more challenging to automate. At the centre of a FOG is the fibre itself, which is available from a wide range of suppliers but which must be selected carefully to meet a host of parameters. These can change depending on the primary design targets, but ideally they will include minimum light losses (preferably below 1 dB/km), polarisation- maintaining properties (to ensure light maintains a linear polarisation as it February/March 2021 | Unmanned Systems Technology Exhaustive testing regimens have been key to the iterative improvements in MEMS devices over the past few years (Courtesy of Physical-Logic) The difference between MEMS IMUs and FOGs comes down largely to the laws of physics and the sizes of their structures
Made with FlippingBook
RkJQdWJsaXNoZXIy MjI2Mzk4