The following are some recent publications from our employees from cybersecurity to air traffic safety.
Cybersecurity in Aerospace Services
As the FAA harmonizes to ICAO standards for aeronautical data and the NAS embraces a cloud infrastructure for collaborative decision making among airlines, air traffic facilities, airports, and industry, the move to streaming services and tapping into data flows needs to pay attention to privacy and cybersecurity. The opportunities for malware to be introduced needs to be actively combated at the instigation of services and continuously monitored. Known vulnerabilities in ADS-B spoofing and Chris Roberts-type interceptions of air-ground data are only the beginning of what is sure to be a growing industry as bad actors attempt to gain admission to US aerospace systems for data interception, malware, ransom, and technology theft. In this paper we explore the Robust Analytics Cybersecurity Framework (RACSF) and the components that can be employed to minimize the threats in the NAS. Topics include the methodology and approaches to building a secure cloud infrastructure, cloud integrity, user verification, validation, data integrity, security monitoring and securing air ground communications.
Buffer Encroachment Trends in the Terminal Airspace across Major US Airports and Impact of Covid-19
Separation assurance is an important facet of the safety guarantee in the the National Airspace System (NAS). Separation requirement for aircraft are defined by horizontal and vertical separation buffers depending on the type of aircraft and type of airspace. The violation of these standards or encroachment into the separation buffers are indicative of reduced safety margins in the airspace. We analyze the buffer encroachment trends in the terminal airspace of 24 major airports in the U.S. over a period of 21 months from April 2019 to December 2020. The analysis measures the duration of encroachments within the terminal airspace for 15-minute epochs and the trends across the airports and their temporal evolution. The period at the beginning of the Covid-19 pandemic in mid-March 2020 and the slow recovery over the remainder of the year is of special interest as it provides a natural experiment for evaluating the impact of unusually low traffic density on aviation operations. Our analysis indicates that the airports show significant variability in the encroachment levels. Some airports like Dallas/Fort Worth International Airport (DFW) and Charlotte Douglas International Airport (CLT) show encroachment levels higher than expected during the reduced traffic period. Furthermore, the trends reveal spikes in encroachment levels during the pre-pandemic period at other airports like Atlanta International Airport (ATL) and Denver International Airport (DEN). The observed trends suggest the utility for real-time measurement of buffer encroachments across airports as a proxy for monitoring safety margin levels and aid the decision-making of different stakeholders.
A Voice Communication-Augmented Simulation Framework for Aircraft Trajectory Simulation
Aircraft operations in the terminal area rely heavily on voice communications between pilots and air traffic controllers. This paper proposes a novel aircraft trajectory simulation framework by guiding the trajectory simulation following the voice command from controllers. Bayesian model selection is used for checking pilot compliances to controller commands with observed trajectories. This framework is named as Voice Communication-Augmented Simulation. The goal of the proposed study is to enable accurate trajectory predictions. The framework can act as a computer assistant for controllers to monitor pilot compliances and ensure safe operations. The proposed method is tested and validated with actual trajectory data from Sherlock Data Warehouse. The tests showed that the proposed framework can accurately simulate and monitor the flight level change of aircraft and update the approach procedure.
Analyzing Controller-Pilot Voice Communications for Safety Analysis
Voice is still the main source of ATC-pilot communications. Those messages offer a large body of airspace situational awareness and operationally-related information that never makes it effectively back into broader NAS use.
Terminal Area Safety Monitoring and Forecasting Presentation
Aviation safety research suggests that safety-related events are
Fatigue and workload, for pilots and controllers
Procedures and training
Crashes and other events usually involve multiple factors
Why not build a system to monitor multiple factors and identify
and predict when one or more risk factors are operating?
Custom IBM Watson Speech-to-text Model for Anomaly Detection using ATC-pilot Voice Communication
In all controlled airspaces, the FAA requires commercial pilots to be in constant contact with and receive instructions from Air Traffic Controllers (ATC). About 80% of all ATC-pilot communications contain at least one error, which result in more than 30% of incidents, flight
level incursions, and runway incursions recorded. Limitations still exist in bridging voice communications into digital exchanges such as broader situational awareness gained on the voice channel, managing time-sensitive exigent non-standard exchanges, and the conveyance
of tone to infer priority. Our innovation detects NAS operational anomalies by uniquely combing with analytical methods our existing TFMData flight information warehouse, live Air Traffic Control (ATC)-Pilot voice communication records, and IBM Watson capabilities such as natural language processing.
Hidden Markov Model based Terminal Area Safety Margin Evaluation Tool (TASMET)
In recent work for ARMD to define a technology roadmap for RSSA, Robust Analytics identified several significant gaps in existing analysis capabilities to support RSSA objectives. One of these gaps, highlighted in meetings with FAA safety experts, was the lack of prognostic risk models that account for NAS infrastructure status. The state of the airspace may contain several major components such as equipment (ALS, ILS, PAPI etc.), weather conditions, and airspace configuration. Unlike the directly observable states of each of the components in the airspace, the safety margin of the entire airspace is inferred, or ‘hidden’. Using a Hidden Markov Model, we relate the sequence of observable states of airspace components, and the sequence of the hidden airspace safety margin level in the Atlanta Hartsfield airport area and
the Atlanta Large TRACON (A80). Our approach enables real-time estimates of NAS risk and can also provide valuable insight into assessments of new technologies and procedures.
DeepLearning Framework for Terminal Airspace Trajectory Prediction and In-time Prognostics
Terminal airspace around an airport is the biggest bottleneck for commercial operations in the National Airspace System (NAS). In order to prognosticate the safety status of the terminal airspace, effective prediction of the airspace evolution is necessary. While there are fixed procedural structures for managing operations at an airport, the confluence of a large number of aircraft and the complex interactions between the pilots and air traffic controllers make it challenging to predict its evolution.