doc. Ing. Jan Roháč, Ph.D.

Theoretical form of navigation solutions in terms of position, velocity, and attitude (PVA) estimates is already well known; however, providing PVA estimates with high accuracy under challenging environmental conditions still plays a key role of current scientific interests. A prime concern is paid to the robustness and accuracy of navigation solutions under real flying conditions where the accuracy can be strongly affected by the varying vibrating, maneuvering impacts as well as environmental impacts and turbulence. Therefore, the area of this PhD study project addresses such problems and will focus on R&D of enhanced integrated solution ensuring improved performance with extended consideration of a sensory assembly. The emphasis will be put onto integration of available navigation sources, e.g. magnetometers, pressure sensors, GNSS positioning, laser scanners, radars etc. to guarantee accuracy even under unfriendly environmental conditions. The navigation solution will be applied on unmanned aerial vehicles (UAVs) whose control systems rely on precise and accurate navigation with a minimal latency. The R&D of such a control system is also going to be a part of this project. This control should then handle situations such as landing of the UAV on a moving platform no matter the environmental conditions and availability of GNSS signal.

The topic deals with a design of human body motion capture system with optimized performance for battery operation. The system will use current technology of inertial sensors aided by magnetometers and GPS receiver. Furthermore, the design will cover a performance study, data processing techniques, tracking algorithms and optimized power consumption with respect to capturing points, data rates, wireless communication and calculation loads.

There are many applications of unmanned vehicles under conditions in which GNSS signals cannot be trusted and/or used at all for navigation purposes. I these situations other aiding systems need to be used, e.g. distance measurement sensors (ultrasound sensors, Lidars), radio-frequency navigation means (wi-fi, UWB), pressure sensors, and in limited way also magnetometers. The topic is focused on a final application including system calibration and heavy experimental testing and verification. The applicability of such a system is broad, for instance for building/vessels/construction inspections, search & rescue missions etc.

UAS as well as RPAS are becoming very popular nowadays and UAS legal usage may be enabled in close future when certain conditions are satisfied. Thus, UAS applicability in emergency, security, and humanitarian services in monitoring the environment and/or search & rescue missions will become an interesting issue for safety and security state services. The flying vehicles have in general limits in their operational range defined by power carried, and of course they will operate in united sky with other flying vehicles. Thus, they need to be reliably controlled no matter whether autonomously or not. For this particular area of applicability, a weather condition should not matter as well as GNSS/GNSS-denied environment or availability and reliability navigation means used. There should be always some potential navigation means but reliability may not be guaranteed. This topic is thus focused on application decision making algorithms utilizing neural networks and/or artificial intelligence to evaluate the level of reliability of available navigation sources and with that respect fusing them. Such an approach will increase an operation safety level for autonomous UAS control and navigation under any weather and diverse navigation-based conditions where reliable and accurate UAS navigation is still required for flying in for instance high-dense urban areas in close distance to for instance building etc. Operating UAS in public areas has always a big challenge in making it safe for surrounding people and assets. The safety of UAS operation is a key parameter which is going to be satisfied via robust modular avionics and control laws supported by accurate navigation.

The topic deals with a design of a navigation system dedicated to terrestrial applications. The emphasis will be put on a primarily inertial system supplemented by aiding system available in GPS denied environment, for instance radars. The aims will cover a performance study, data processing techniques, tracking algorithms, data validation and adaptive data fusion with optimized performance for dynamically changing conditions and environment.

GNSS based navigation is a common approach to obtain position and velocity of a navigated object; nevertheless, a GNSS signal is relatively weak and thus is vulnerable to disruption as well as in dense urban areas may be unreliable and/or unavailable at all. With this respect there is a need to seek for potential other sources of navigation suitable signals, i.e. signals of opportunity (SoO). Such navigation with incorporated SoO will be able to calculate its position by making use of the hundreds of different signals that are all around. The SoO based navigation system will learn from signals that are initially unidentified to build a local RF signal fingerprint, i.e. identify and localize sources of SoO, when a GNSS signal is available. By incorporating the local RF signal fingerprint with such a wide range of signals into navigation algorithms will ensure the navigation accuracy even under GNSS denied conditions and/or under unreliable/jammed GNSS signals. The system will be primarily using an inertial measurement unit and GNSS, but incorporating SoO will provide superior performance to GNSS and further increase the accuracy and robustness as well as safety of autonomously operated vehicles relying on accurate positioning and control. Of course, this principle can be used to any kind of vehicle when navigation is required. Using SoO does not require to build a new infrastructure but it uses the one already in place. A major advantage of such a system is its ability to function in places where GNSS is unable to reach, e.g. dense urban areas and inside buildings, when the RF signal fingerprint is available. Such an approach provides a cost-effective solution with a wide variety of different indoor/outdoor applications.

Navigation in general is challenging in cities where GNSS signal may be often blocked or week and thus cannot provide sufficient accuracy for localization. Since this localization approach has its limits in cities it is necessary to seek for other possibilities to ensure the navigation capability and its robustness. The topic addresses this navigation problem with using multi-agent approach in a broad network of public services and cooperating devices. The navigation will consist of functionality to share data within a network of devices in a local area and use positions of other devices if known for localization of those which do not have a direct possibility via for instance GNSS or other means. Localization will be performed based on relative positioning between a particular device of interest and surrounding devices with known position. Such knowledge of position might be based on for instance GNSS signal or mediated from the network of cooperating devices. This intermediation might have endless deepness/levels but with increasing number of levels of intermediation accuracy decreases. This principle will further take into account that an individual device position may suffer from large uncertainties since GNSS signal low reliability in high-dense urban areas or the position has been already intermediated; nevertheless, when more relative positions even with higher uncertainties are fused together the resultant accuracy will improve. Moreover, it will be studied if data processing incorporating failure resistive algorithms can isolate intentional or unintentional incorrect data provided by the users. This leads to involve data validation, probability theory and artificial intelligence to the localization algorithms and cloud-based computation of potential uncertainties in the network. The approach requires a cooperation of large number of agents (users/devices) sharing the data via Bluetooth/Wi-Fi and mobile networks. The relative positioning will be based on RSS (Received Signal Strength) evaluation performed in a device of interest. This kind of localization can be used for personal and/or terrestrial vehicles navigation in cities with high density of potential cooperating devices. It can be further used for drones positioning, but the emphasis in this case is put onto personal and terrestrial vehicle navigation.