Aviation safety: a framework to analyse airspace infringements involving general aviation flights

Abstract

The increasing number of commercial flights, e.g. a scheduled passenger flight operated by an airline, requires that all air aviation operations must follow the airspace rules and procedures in order to enhance safety. Since commercial flights comprise the majority of civilian flights, most of a country’s airspace is designed to meet its needs. As a consequence, airspace design and procedures primarily meet the needs of such flights. For example, controlled airspace is under the jurisdiction of the Airspace Navigation Service Provider (ANSP) and thus, the Air Traffic Controller (ATCO) is responsible for the management of flights within the airspace including authorisation for entry and exit. There are, however, situations in which aircraft enter controlled airspace without authorisation of the ATCO. When this happens, a safety incident known as an airspace infringement (AI) is recorded. Such incidents can cause safety as well as other air traffic management problems, such as delays, with the worst case being a mid-air collision. According to European statistics, approximately 5% of reported AIs in European airspace led to a loss of separation between 2010 and 2014. A distinctive characteristic of AIs is that they mostly involve general aviation (GA) flights that significantly differ from commercial flights. An immediate consequence of this is that current reporting, investigation and analysis schemes are designed for commercial aviation and thus, incidents involving GA are inadequately reported in the safety database of ANSPs and national aviation authorities. This unsurprising given that on average approximately 20% of reported AIs in Europe were not analysed between 2010 and 2014. Furthermore, not all GA aircraft are adequately detected by the air traffic service infrastructure and thus, AIs might be underreported. GA in itself represents a diverse group of airspace users and has unique relationships with aviation stakeholders at the national and European level which are not well described in the literature. Two seminal studies of AIs that focused on GA were conducted in the last decade by the European Organisation for the Safety of Air Navigation (EUROCONTROL) and Civil Aviation Authority of United Kingdom (UK CAA). These studies provided a relatively good overview of the context in which AIs occurred in European airspace in the period 2003-2008. Whilst important contributory factors to AIs were determined, they did not relate to a GA pilot’s decision making, the flying style of each GA aircraft or the use of emerging technologies. Such factors can be found in the literature relating to human factors and aviation psychology. Nonetheless, the factors of these AI studies were not found in reported AI incidents, which are the key data for any safety analysis. Despite efforts to mitigate their occurrence at both national and European level, AIs continue to occur both in controlled and restricted airspace. Although a small proportion of AIs results in a loss of prescribed separation between aircraft, in the absence of evidence to the contrary, AIs pose the risk of a mid-air collision in controlled and restricted airspace. GA remains the key contributor to these incidents whilst it faces technological and regulatory changes that can affect their operations and subsequently, the occurrence of AIs. Significant improvements have also been made to the safety databases of aviation stakeholders. However, according to recent European safety reports, there is a need for both a systematic reporting and safety analysis to meet the increasing size of the safety databases. Therefore, this thesis addresses these current limitations and proposes a holistic framework to analyse AIs at the national level to enhance safety. This framework utilises both the developed taxonomy of contributory factors for GA and the developed framework that systematically analyses reported AIs. This thesis also reviews the GA flying activity including the operations and a pilot’s performance. The developments of this thesis are summarised below. GA is extensively described from a review of various documents and interviews with the GA pilots conducted in this thesis. The latter reveals the manner by which recreational GA pilots currently plan and navigate as well as the information they need to do so. This description can also be used by the stakeholders to identify the needs of the GA pilots and to review airspace rules and procedures to meet these needs. With respect to AIs, this thesis develops and validates a novel and bespoke taxonomy of contributory factors of AIs, referred to as GA-Contributory factors to AIs (GA-Saf(ArI)) for GA flights. The taxonomy is derived from an extensive literature review of GA pilots’ performance, past studies of AIs and the rigorous analysis of reported AIs and semi-structured interviews with GA pilots. The taxonomy consists of 145 contributory factors and 52% of these factors were derived solely from the interviews, highlighting the importance of collecting data directly from GA pilots. Amongst the findings is that the flight-route decision of GA pilots to fly very close to controlled airspace, mostly over capital cities, increases the likelihood of an AI. In such flights, the aircraft can stray into controlled airspace when the pilot fails to notice the change of the aircraft’s heading due to the wind. The taxonomy GA-Saf(ArI) can be used by the stakeholders to analyse AIs as well as other incidents involving GA flights. This thesis also develops, tests and validates a framework, referred to as Proactive Safety Analysis of Reported AIs (Pro-Saf(Ari)). This framework overcomes current limitations through a systematic and speedy analysis of reported incidents. It assesses the quality of the data, extensively studies the incident’s contextual and contributory factors and applies statistical analysis. Given the current database’s structures, the success of the analysis depends on the narrative. The more detail the narrative is, the more likely the contextual and contributory factors to be determined. The uniqueness of AIs is that most AIs are reported by the ATCO than the infringing pilot. Whilst their detailed narrative is useful, it fails to report information on what happened inside the aircraft, such as the flight-route decision making of pilots. The framework is applicable to databases of different quality, size and origin, i.e. ATCO’s and pilot’s reports, and thus, it can be used by the stakeholders. Finally, this thesis develops a holistic framework, referred to as Safety analysis of AIs at the national level (Pro-Saf(ArI)N). The Pro-Saf(ArI)N framework offers a systematic approach to collect, investigate and analyse the AIs reported by both the ATCO and GA pilot. It utilises the taxonomy GA-Saf(ArI) to investigate the reported incidents and the framework Pro-Saf(Ari) to statistically analyse the contextual and contributory factors. Such a holistic framework allows the stakeholders to monitor AIs, assess the safety risk of a mid-air collision caused by an AI and design a mitigation action plan to prevent AIs from occurring at national level. A quantitative method is applied to measure the efficacy of a mitigation action. From the application of the framework, a series of mitigation actions are proposed. Furthermore, improvements of the Safety Management System of the ANSP are proposed including the improvement of the reporting scheme.