Separation assurance, self-separation, and collision avoidance
In Unmanned Aircraft Systems (UAS), the ability of an aircraft to maintain a safe distance from the other aircraft without any input or direction of human regulators, including air traffic control, is known as self-separation. The system consists of a number of coupled factors such as communication performance, route planning, airspace capacity management, overtaking rules, navigation and surveillance, and collision avoidance strategies (Lee & Mueller, 2013). Each of these factors can lead to dynamic changes in the whole air traffic. On its part, the airborne separation assurance system (separation assistance) is the system in the UAS that lets the crew keep the aircraft apart from others and provides flight information regarding the surrounding traffic.
Collision avoidance is an automated system in UAS that autonomously senses obstacles during the aircraft operation and helps it to avoid colliding with them in its mission by using optical sensors as well as intelligent software systems. In drone technology, for instance, the FlytCAS is an intelligent software system whose function is to enable collision avoidance on commercial drones (Narkawicz et al., 2017). Collision avoidance systems have the ability to sense objects in real-time and avoid imminent collisions between the aircraft and the objects on its path, including in complex environments.
Sensing, conflict detection, and avoidance maneuvering in DAA architecture
The autonomous operation of UAS aircraft is highly complex and aims a providing operation approval, sensing and detection, and avoidance maneuvering functions when in operation. Various developers have come up with systems designed to achieve these objectives in UAS operations. The Detect and Avoid Alerting Logic for Unmanned Systems (DAIDALUS) is a relatively recent implementation and the basis for ICAROUS software architecture and whose purpose is to detect and avoid (Davies & Wu, 2018). The system consists of self-separation and alerting algorithms that have the capability to provide situational awareness to UAS remote pilots in the form of guidance in maneuvering. The aim is to aid in maintaining and regaining clear separation for unmanned aircraft. The DAIDALUS system provides a mission agnostic as well as highly configurable application program interface to allow the integration of diverse mission platforms and applications.
Onboard the unmanned aircraft, the function of detection and avoidance can be achieved using the DAIDALUS system. The system is a software library whose purpose is to implement a DAA concept for the integration of the aircraft into the National Airspace System (NAS). DAIDALUS architecture is designed in such a way that it uses a parametric volume that helps the unmanned aircraft pairs occupying the same volume be considered to be in the same violation (Cisco, 2007). The algorithms in the system predict the potential well-clear violation within a specific look-ahead time. They make clear assumptions of the non-maneuvering trajectories and determine the instantaneous well-clear status between a pair of unmanned aircraft.
Cooperative sensing technology
A mature cooperative system is known as the Traffic Alert and Collision Avoidance System (TCAS), which has been widely used in manned airplanes. In TCAS, a transponder is used to send data during the detecting process. To prevent collisions, an aircraft with a TCAS can communicate with other aircraft that have TCASs as well. The most recent TCAS version can identify objects up to 160 km away, while the improved version can detect objects up to 129 km away (Yu & Zhang, 2015). TCAS is suitable for both the Visual Meteorological Condition (VMC) and Instrument Meteorological Condition (IMC).
First, it is able to provide accurate as well as reliable information on variables during the mission. Secondly, it uses GPS, which is well-proven communication technology. Third, it has a flexible structure that enables ease of implementation and incorporation of other techniques in the future. Nevertheless, it has some limitations, such as ineffectiveness in detecting and avoiding ground-based obstacles such as towers, power lines, terrain features, and others. Regarding the TCAS sensor, it provides information regarding the range altitude (Cook et al., 2015). It is favorable in use for getting information regarding Visual Meteorological Condition (VMC) and Instrument Meteorological Condition (IMC), but is not favorable in terms of size, weight, and power (SWAP), and is expensive. It is remarkable that this device is well proven and widely used.
Non-cooperative sensing technology
LIDAR estimates distance by lighting an object with a laser to examine the reflected light. LIDAR has been widely studied in UASs as a remote sensing technology to portray terrain through the related collections of spatially dispersed points with precise coordinate triples. In the Robotics Institute at Carnegie Mellon University, a LIDAR system is employed to lower the number of false-positive rates and provide the distance of the invading aircraft (Terwilliger et al., 2017). Tree height assessment and digital terrain model improvement are tested for assessment in a unique mini-UAS-borne LIDAR system with flexibility.
In this instance, the LIDAR is created using the Ibeo Lux laser scanner. LIDAR systems typically have detection ranges between 200 m and 3 km. The advantage of LIDAR is that it can identify objects as small as 5 mm in diameter and as large as skyscrapers while recognizing non-perpendicular areas at high resolution. Similar to that, it is extremely adjustable and suitable for various atmospheric conditions. The Field of View (FOV) is constrained, which is a drawback. Regarding the sensor used, LIDAR assesses range and is applicable for VMC and IMC, yet it is not favorable in terms of SWAP and cost (Yu & Zhang, 2015). One of its primary benefits is the easy configuration required for its proper functioning.
Path planning approach
The course of UASs has recently been planned using randomized algorithms. The configuration space of the UAS is typically sampled randomly to develop a one-dimensional graph. A tree of possible trajectories is formed online by growing branches toward randomly generated target locations after conducting a stochastic search across the body-centered reference frame. An incremental roadmap construction algorithm can handle changing impediments and system dynamics (Cook et al., 2015). Then, for path planning in the context of both stationary and moving obstacles, a Rapidly Exploring Random Tree (RRT) search-based algorithm is used. RRT-based path planning is a proposed strategy for preventing collisions between miniature air vehicles (MAVs). The RRT technique is used to find a collision-free path using a depth map that represents the range and direction of obstructions. If a viable path exists, which signals that it is stochastically complete, the probability of discovering a route from origin to destination can converge to one under the right technical circumstances. An RRT-based path planner incorporates the probability constraint model to account for ambiguity to cope with uncertain dynamic impediments.
In the study, the authors addressed the issues with RRT regarding the great and random variety of results, the low speed of convergence, and deviation. To find the solution for them, a modification to RRT was developed using the cyclic altering iteration search method (Shi et al., 2020). The improvement was evaluated with the simulation to examine if the increase in quality is evident. It showed that the problems addressed in the study were solved successfully, which led to shorter path length, and the path is smoother. The study concludes that all algorithms, whether they are enhanced Rapidly-Exploring Random Tree algorithms, particle swarm algorithms, or other types, are model-driven and have specific constraints. To eventually advance the technology of intelligent cars, further research is required to integrate data-driven course planning and obstacle avoidance.
References
Cisco (2007). Antenna Patterns and Their Meaning. Web.
Cook, S. P., Brooks, D., Cole, R., Hackenberg, D. & Raska, V. (2015). Defining Well Clear for Unmanned Aircraft Systems. AIAA SciTech Forum. Web.
Davies, J. T. & Wu, M. G. (2018). Comparative Analysis of ACAS-Xu and DAIDALUS Detect-and-Avoid Systems.
Lee, S. M. & Mueller, E. R. (2013). A Systems-Based Approach to Functional Decomposition
and Allocation for Developing UAS Separation Assurance Concepts. AIAA AVIATION Forum.
Narkawicz, A., Munoz, C. & Consiglio, M. (2017). A Path Planning Algorithm to Enable Well-Clear Low Altitude UAS Operation Beyond Visual Line of Sight. Twelfth USA/Europe Air Traffic Management Research and Development Seminar.
Shi, Y., Li, Q., Bu, S., Yang, J., & Zhu, L. (2020). Research on intelligent vehicle path planning based on rapidly-exploring random tree. Mathematical Problems in Engineering, 2020. Web.
Terwilliger, B., Ison, D. C. & Robbins, J. (2017). Small Unmanned Aircraft Systems Guide. Aviation Supplies & Academics, Inc.
Yu, X. & Zhang, Y. (2015). Sense and avoid technologies with applications to unmanned aircraft systems: Review and prospects. Progress in Aerospace Sciences 74, 152-166. Web.