In recent years, the overall performances of inertial Micro-Electro Mechanical Sensors (MEMSs) exhibited substantial improvements to values very close or similar to so-called tactical-grade sensors. However, due to their high costs, numerous researchers are currently focusing on the performance enhancement of cheap consumer-grade MEMS inertial sensors for all those applications (as an example, small unmanned aerial vehicles, UAVs), where cost effectiveness is a relevant request; the use of redundancy proves to be a feasible method for this purpose. In this regard, the authors propose, hereinafter, a suitable strategy aimed at fusing raw measurements provided by multiple inertial sensors mounted on a 3D-printed structure. In particular, accelerations and angular rates measured by the sensors are averaged according to weights associated with the results of an Allan variance approach; the lower the noise figure of the sensors, the greater their weight on the final averaged values. On the other hand, possible effects on the measurements due to the use of a 3D structure in reinforced ONYX (a material capable of providing better mechanical specifications for avionic applications with respect to other solutions for additive manufacturing) were evaluated. The performance of a prototype implementing the considered strategy is compared with that of a tactical-grade inertial measurement unit in stationary conditions, exhibiting differences as low as 0.3 degrees in heading measurements. Moreover, the reinforced ONYX structure does not significantly affect the measured values in terms of both thermal and magnetic field while assuring better mechanical characteristics with respect to other 3D printing materials, thanks to a tensile strength of about 250 MPa and a specific stacking sequence of continuous fibers. Finally, a test conducted on an actual UAV highlights performance very close to that of a reference unit, with root-mean-square error in heading measurements as low as 0.3 degrees in observation intervals up to 140 s.

In this paper, the authors investigate the possibility of applying artificial intelligence algorithms to the outputs of a low-cost Kalman filter-based navigation solution in order to achieve performance similar to that of high-end MEMS inertial sensors. To further improve the results of the prototype and simultaneously lighten filter requirements, different AI models are compared in this paper to determine their performance in terms of complexity and accuracy. By overcoming some known limitations (e.g., sensitivity on the dimension of input data from inertial sensors) and starting from Kalman filter applications (whose raw noise parameter estimates were obtained from a simple analysis of sensor specifications), such a solution presents an intermediate behavior compared to the current state of the art. It allows the exploitation of the power of AI models. Different Neural Network models have been taken into account and compared in terms of measurement accuracy and a number of model parameters; in particular, Dense, 1-Dimension Convolutional, and Long Short Term Memory Neural networks. As can be excepted, the higher the NN complexity, the higher the measurement accuracy; the models’ performance has been assessed by means of the root-mean-square error (RMSE) between the target and predicted values of all the navigation parameters.

View Video Presentation: https://doi.org/10.2514/6.2023-2705.vid This paper describes an innovative method to measure the heading angle of a platform by exploiting the polarization state of Sunlight. Measuring the polarization state of Sunlight in a certain celestial region, the Rayleigh Single Scattering model can be used to identify the Sun meridian/anti-meridian in a polarized image to support navigation systems. A polarimetric camera is used to acquire several images containing the Sun meridian/anti-meridian. Once the Sun azimuth and attitude data are acquired at the current time and location, the position of the Sun meridian/anti-meridian in the local reference frame can be compared with its position expressed in the camera reference frame by exploiting inertial sensors. In this way, an accurate heading angle measure can be performed to improve the navigation performance of autonomous vehicles. Several tests have been carried out: (i) ground tests were performed to evaluate the proposed method by using a polarimetric camera and a certified inertial unit on a tripod; (ii) flight tests were performed to assess the achieved performance installing the polarimetric camera on-board a drone. As shown by the obtained results, the proposed method ensures accurate heading angle estimates by exploiting the polarization vector information that can be used in an autonomous integrated navigation system when satellite navigation is not available in nominal conditions.

The growing employment of Unmanned Aerial Systems (UASs, commonly referred to as drones) in civilian and military environments, along with the increasing accuracy of avionics components, has led to the definition of more stringent drone-based mission requirements. In this sense, the following work summarizes a series of Ph.D. research activities aimed at improving UAS performance under different aspects, in order to enhance their use in a variety of application contexts and environmental conditions while ensuring a satisfactory level of accuracy.