Vittorio Di Vito

Automatic Collision Avoidance System: Design, development and flight tests

This paper presents a fully Automatic Collision Avoidance System (ACAS) for unmanned aerial vehicles. This system has been developed by the Italian Aerospace Research Center (CIRA) in collaboration with the department of Aerospace Engineering of the University of Naples “Federico II”, in the framework of the national funded research project TECVOL (Technologies for the Autonomous Flight). The proposed system is comprised of two subsystems: the Obstacle Detection and Tracking subsystem, which permits to reveal flying intruders in a selected field of regard and to estimate their motion; the Collision Avoidance subsystem, which provides conflict detection and resolution capabilities, addressed in a 3D environment using information about current position and instantaneous speed vectors. The effectiveness of the system has been demonstrated during a flight test campaign, where proper conflict scenarios have been considered. In fact, the proposed ACAS setup was installed onboard a very light aircraft named FLARE (Flight Laboratory for Aeronautical Research), which has been customized with automatic flight capabilities. System architecture and the developed algorithms are described, then some results obtained from the flight test campaign are presented and discussed which demonstrate the reliability and the efficiency of the developed system.

On-line trajectory generation for autonomous unmanned vehicles in the presence of no-fly zones

In this article, an algorithm for three-dimensional path generation and tracking for unmanned air vehicles in the presence of no-fly zones is proposed. The algorithm is based on a local optimization procedure aimed to find the shortest path between the waypoints in compliance with all path constraints. Vehicle structural and envelope limitations are accounted for by simple geometric constraints such as minimum curvature radius and flight path angle limitations, while no-fly zones are defined as cylindrical objects with infinite altitude. The algorithm is simple and it has a limited computational burden, at most quadratic with the number of zones to avoid. This makes the algorithm very suitable for real-time applications even in case of a high number of forbidden zones. Algorithm effectiveness has been demonstrated by means of numerical simulations in scenarios including the presence of no-fly zones not known before flight (for instance, in the case of sudden changes of weather conditions and/or detection of new fixed obstacles).

Flight Testing of a Fully Adaptive Algorithm for Autonomous Fixed Wing Aircrafts Landing

This paper presents the flight test results of a fully adaptive algorithm for autonomous fixed wing aircrafts landing developed by CIRA, the Italian Aerospace Research Centre. The algorithm is designed and implemented in the framework of a complete autonomous guidance system, worked out by CIRA, able to allow autonomous way-points navigation, autonomous landing and autonomous collision avoidance for fixed wing aircrafts. The algorithm presented in the paper is designed to perform a fully adaptive autonomous landing starting from any point of the three dimensional space, based on the use of the DGPS/AHRS technology. Main features of the autolanding system based on the implementation of the proposed algorithm are: on line landing trajectory re-planning, fully autonomy from pilot inputs, weakly instrumented landing runway, ability to land starting from any point in the space and autonomous management of failures and/or adverse atmospheric conditions. The flight tests have been conducted at an airfield in Caserta, in the south of Italy, close the CIRA. The paper is structured into several paragraphs describing the algorithm designed for the autolanding maneuver, the control system architecture and the methodologies developed in order to safely manage the possible presence of failures and/or unfavorable weather conditions, the preliminary results of the real time validation with hardware in the loop simulation and, finally, the performances achieved by using the CIRA experimental flying platform, with reference to the real flight experiments.

An Overview on Systems and Algorithms for On-Board 3D/4D Trajectory Management

In recent years, significant emphasis has been devoted to progresses in the field of on-board trajectory generation and tracking for aircraft guidance along specified waypoints. Research in this area was mainly driven by the request of improved on-board autonomy for both unmanned aerial systems and manned civil aircrafts. Several algorithms and systems have been developed and patented in this field, addressing both 3D waypoints, which are specified in terms of their three dimensional coordinates and, more recently, 4D waypoints, which also include a required time to arrival. Of course, patents proposed for 3D/4D trajectory management systems are characterized by their key features (use of 3D or 4D constraints, of different waypoint capture concepts, and so on) and by different applications and algorithms. In this paper, a review is performed of the most relevant patents presented in recent years in this field. This review considers the various problems, emphasizing main advantages and weaknesses of the proposed solutions and suggesting some possible future developments.

In-flight performance analysis of a non-cooperative radar-based sense and avoid system

This paper discusses the outcome of flight experiments relevant to non-cooperative sense and avoid that were carried out in the framework of a collaboration between the Italian Aerospace Research Center and the University of Naples “Federico II”. Within the project, aimed at full autonomy for medium/large Unmanned Aircraft Systems, an integrated radar/electro-optical system configuration was adopted, and real-time data fusion/automatic decision-making algorithms were developed to estimate intruders’ motion, and generate and follow proper escape trajectories. All the systems were designed to guarantee collision avoidance without any dependence on the command and control link. The hardware/software sense and avoid system was installed onboard an optionally piloted flying laboratory of Very Light Aircraft category, and a single intruder aircraft of the same class was considered for avoidance experiments. In particular, the paper focuses on the results from radar-based autonomous avoidance flight tests. Performance is analyzed in terms of situational awareness before and during the maneuver, and effectiveness in performing safe avoidance maneuvers while generating smooth commands and minimizing the deviation from nominal trajectory. Statistics relevant to several encounters show that in spite of a small number of valid sensor measurements, the tracking algorithm is able to keep a satisfying level of situational awareness, thus enabling safe avoidance. However, the limits deriving from coarse radar accuracy, low scan rate, and interaction between mechanical scanning and aircraft flight dynamics are also pointed out. While these limits did not impact avoidance safety in the considered tests due to the favorable operating environment in terms of time to collision, radar detection range, and ownship maneuverability, they confirm the need of improved sensing concepts in more challenging avoidance scenarios, such as adoption of radar/electro-optical data fusion and electronic scanning technologies.

Automatic Guidance through 4D Waypoints with time and spatial margins

This paper presents a new algorithm for 4D Automatic Guidance, enabling the automatic capture of 4D waypoints (i.e. 3D points in the space with a requested time of arrival), while satisfying navigation constraints coming from both Air Traffic Management (i.e. time and space constraints on the waypoints to reach) and vehicle performance limitations. The proposed guidance strategy mainly relies on the continuous re-generation of the geometric reference trajectory and, in case this is not sufficient to guarantee the requested time of arrival, airspeed reference is adjusted to compensate for the difference between estimated and requested time of arrival of waypoint. The effectiveness of the proposed approach has been demonstrated by means of numerical simulations, including vehicle model and simplified autopilot/autothrust system that ensure tracking of the reference 4D trajectory. This paper was developed in the framework of 4DCoGC, a research project funded by European Commission under the Seventh Framework Programme whose objective is to address the aircraft 4D guidance and control principle.

Real-Time Validation of an ADS-B Based Aircraft Conflict Detection System

In the framework of future surveillance systems, ADS-B devices are assuming a primary role. They are expected to be widely implemented on-board of vehicles, so the possibility of using ADS-B data to support the safety of the flight is worth to be investigated. This paper describes the real time validation of an ADS-B based application for conflict detection algorithms in order to support future self-separation as well as collision avoidance systems. In particular, the paper first presents an overview of the proposed system architecture including a brief description of the software modules and, then, provides a description of the hardware-in-the-loop architecture that has been used for the validation of the system. Finally, in the paper the results of some real-time simulation tests are described and discussed. The results reported in the paper emphasize that the proposed conflict detection system, based on the use of ADS-B data, is suitable for real-time implementation and, therefore, can be considered for the implementation on-board vehicles for on-line application during the flight.

An advanced 3D algorithm for automatic separation assurance systems

Hazardous collision situations among multiple aircraft could arise in future air transport system more often than today, since traffic density is predicted to hugely increase. An Automatic Conflict Resolution Algorithm is presented in this paper, which can provide automatic aircraft separation assurance for next generation air traffic control systems. Once a conflict arising with one or more aircraft has been detected, the proposed algorithm computes a safe flight trajectory to maintain separation with other traffic. The algorithm has been conceived for installation on-board autonomous aircraft because it performs required decision-making process for conflict detection and generates commands directly to the autopilot for conflict resolution. The algorithm is not rule-based but uses an efficient analytical solution to detect the conflicts, thus resulting suitable for real-time applications. Moreover, the proposed strategy for generating the conflict resolution maneuver minimizes the deviation from the original trajectory. In addition to algorithm description, some numerical simulations with challenging scenarios are presented in the paper, showing algorithm ability to solve a conflict situation in a self-organizing system including several aircraft equipped with the same proposed algorithm.