Scott Leishman's Unmanned blog

Air Traffic Control Human Factors with Drones

Scott E. Leishman

Embry-Riddle Aeronautical University – Worldwide

ASCI Graduate Capstone Project

Submitted to the Worldwide Campus

In Partial Fulfillment of the Requirements of the Degree of

Master of Aeronautical Science

July, 2018

Abstract

An area of the National Airspace System that is deeply impacted by volume of aircraft is air traffic control (ATC). As the volume increases, the demand for a controller’s workload does as well. Research has demonstrated without adequate personnel, sufficient automation, and lack of accepted technology, ATC workload can and will lead to an increase in delays, incidents, and mishaps. This project illustrates the various human factors that affect ATC, and how the addition of unmanned aircraft impacts National Airspace system (NAS). The project found correlation between controller workload and the incorporation of unmanned aircraft systems. The correlation indicated that controllers can become overloaded, due to the unique nature of unmanned aircraft, and are susceptible to create loss of separation events. This project also details the economic impact unmanned aircraft systems (UAS) have on the United States economy. Key discussions throughout this research inquiry are directed at understanding what automation is needed, what technology is viewed as adequate, and how to decrease workload while maintaining efficiency in the NAS. With this research inquiry, alternative controlling techniques are reviewed, to include creating a separate airspace system that could work in unison with unmanned aircraft and manned aircraft.

Keywords: Air traffic control, National Airspace system, workload, stress management, unmanned aircraft systems, safety management systems, Federal Aviation Administration

Air Traffic Control Human Factors with Drones

Proposal

Objective. The research proposed for this project will examine the effects of having an increased workload with a diminishing number of qualified air traffic controllers. More specifically, this research will be directed towards how air traffic controllers will manage unmanned aircraft, an area in aviation that does not have extensive research documented. There are a variety of issues that need to be addressed to fully implement unmanned aircraft in the NAS. The findings from this research will aid future lawmakers and technology developers to create systems that can work effectively together with the human-machine interface.

Scope. Fully certified professional controllers (CPC) are sitting at a 30-year low (Gilbert, 2016), with only 10,667 CPC within the United States. This decrease began in 2011 and has caused staffing of controllers to decrease by 10% ("2018-2027 Air Traffic Control Workforce Plan", 2018). With the total time for a controller to get qualified at busier facilities taking on average approximately 2 years, should this trend continue, there will not be adequate personnel to handle the influx of unmanned aircraft. In a study performed by the U.S. Department of transportation, UAS are expected to grow from a few hundred to approximately 10,000, with over 90 percent of these vehicles categorized as Nano, Micro, or Small UAS for Federal Agency use. Alternatively, as of 2016 there have been over 626,000 hobbyists who have registered to operate a non-commercial UAS. This number has shown the potential to increase to 4.5 million over the next 20 years (Glaneuski, 2013).

This becomes problematic for a couple of reasons, with one being hiring plans for future air traffic controllers versus retirements. Other issues include legislation governing UAS and problems arising from inadequate technological updates to the NAS. The NAS has fallen behind on incorporating important technology, this technology is important for full implementation of UAS platforms. Having technology such as the low-altitude tracking and avoidance system (LATAS) or unmanned traffic management (UTM) would greatly enhance an air traffic controllers’ ability to manage the airspace in which UAS operate. Without conducting research into human factors issues related to ATC, and understanding how this affected by UAS, it would be difficult to assess how quickly UAS could be implemented and utilized. This problem will only become larger without countermeasures, sufficient automation, updates to technology, legislation, or a proactive plan (Sondgeroth, 2018).

Methodology. To understand the role ATC human factors plays regarding UAS in the NAS, a quantitative research approach will be used to evaluate issues affecting human factors, staffing levels, and technology made available to controllers. Using this same approach, the project will demonstrate components of the SHEL model, and how each of those interfaces are impacted. The data used for this project will be gathered from previous human factor studies involving air traffic controllers. Limitations to this project include not enough real-world data involving unmanned aircraft. The data for this project will be obtained from research studies that measure workload and performance of air traffic controllers. Task demand and workload play an important role on safety and efficiency of air traffic. This project will use ATC human in the loop (HITL) simulation to investigate task demand variation on workload and performance.

Because air traffic volume is expected to increase, and due to the lack of adequately fully qualified CPC, legislation and technology need to be cognizant of future trends. To ensure the NAS is not compromised, this project will address and outline key issues and detail potential solutions with the observations made during the research process.

Program Outcomes that will be addressed in this project:
Program Outcome #1

Students will be able to apply the fundamentals of air transportation as part of a global, multimodal transportation system, including the technological, social, environmental, and political aspects of the system to examine, compare, analyze and recommend conclusions.

An investigation into how the air traffic control system is affected will be presented for this project using the SHEL (Software, Hardware, Environment, and Live-ware) model, which details human factors issues. The social aspect of the ATC system will be addressed by evaluating how different components of that environment get affected by staffing shortages, and changes in volume. Using the SHEL model, the social aspect can be addressed under live-ware and environment.

The improvements the NAS is experiencing through NexGen will help to illustrate areas of technology affecting ATC as well as the implementation of UAS. Perhaps the most critical component of implementing so many new aircraft into the NAS, is the technology that supports this type of aircraft is being incorporated. Going from ground-based navigation to satellite-based navigation will enhance an air traffic controller’s ability to accurately identify aircraft. Other countries have incorporated drone technology into their NAS (global, multimodal) which this project will review to see if there are certain political or technological advances these other countries have made to make it possible for UAS to operate, and how they adequately staff their ATC workforce. The environmental impacts of this implementation will be reviewed, as delays due to staffing could become costly and problematic for the air transportation industry. This project is seeking to find gaps/needs in technology, legislature, training, and correct headcount for amount of air traffic expected over the next 20 years, and to provide future recommendations.

Program Outcome #2

The student will be able to identify and apply appropriate statistical analysis, to include techniques in data collection, review, critique, interpretation and inference in the aviation and aerospace industry.

Air traffic control shortages stem from the Professional Air Traffic Controllers Organization (PATCO) strike that occurred during President Reagan’s administration in 1981. This strike and the onslaught of layoffs has since hindered the FAA in being able to adequately staff for traffic which, with the impending changes expected to occur with the implementation of UAS, is problematic. Data is available on staffing, there are also several studies conducted that highlight air traffic control workload. This data will be collected and then utilized to compare and analyze human factor related issues on workload, performance, and efficiency. Using this study and data points from the FAA, this data will then be analyzed using an ANOVA test to infer if the hypothesis that was proposed will be accepted or rejected. The dependent variable for this project will be air traffic control performance and the independent variable for this project will be variance in traffic density. To construct this statistical analysis, traffic load index will be the primary factor measured. The study to be used for this will be one performed by Ames Research center. The study itself simulates air traffic control workload when UAS are present. The study itself utilizes controllers with prior experience and investigates UAS impact on performance and efficiency of the air traffic controller.

Program Outcome #3

The student will be able across all subjects to use the fundamentals of human factors in all aspects of the aviation and aerospace industry, including unsafe acts, attitudes, errors, human behavior, and human limitations as they relate to the aviators adaption to the aviation environment to reach conclusions.

As mentioned in program outcome # 1, the SHEL model will be used to analyze human factors within air traffic control. Secondary information will be utilized to analyze the same components of SHELL, but with UAS and their operation in the NAS. Having the ability to do this will help to illustrate what unsafe acts could be generated based on the decrease of fully qualified controllers, and the increase in expected air traffic. Human behavior will become a factor going forward, as controller reactions will determine procedures that need to be sought in the legislative process. The project will address attitudes about how UAS operations have stricter regulations than their manned counterparts. Mainly, this project will seek to find resolution in an area of aviation that is still relatively new and potential issues that arise from these sorts of operations. With UAS operations, certain human limitations take place, as many components of UAS operations are autonomous. To mitigate these issues, this project will look specifically at technology that aids UAS operators as well as air traffic controllers. This will help to unveil areas of human limitation and how conclusions can be made to overcome and adapt to these limitations.

Program Outcome #4

The student will be able to develop and/or apply current aviation and industry related research methods, including problem identification, hypothesis formulation, and interpretation of findings to present as solutions in the investigation of aviation/aerospace related topic.

A quantitative study will be conducted and the results identified to accept or reject the hypothesis that there is a statistically significant difference for air traffic controller workload when UAS are incorporated. The study will examine measures on workload based off of performance and efficiency. Controllers with previous experience in that control sector will be utilized, and will be tested with the number of UAS present in each simulation. The UAS will be given a simulated verbal response delay and will vary in speed (slow, mixed, and fast). The independent variable for this project will consist of UAS, while the dependent variable will be the air traffic controllers. As mentioned in program outcome #2 the study that is used will compare the effects of different levels of task load on subjective measures and performance. The key tasks in the study were conflict detection, conflict resolution, arrival metering (schedule conformance), and monitoring the automation when it was completing these tasks. The dependent variable will be a controller’s workload, what the workload looks like under diminished automation versus what it looks like when new technology is incorporated.

To further illustrate the dependent variable, workload will be measured on standard operating procedures used by air traffic controllers. This analysis will build upon previous research and what was discovered about automation and workload relationships. As mentioned in program outcome #2, an ANOVA test will be utilized for the findings and will be used to make interpretations. These interpretations will help to provide insight into future policymaking and aid technology manufacturers develop a plan for the NAS. Secondary findings will illustrate weaknesses and strengths of each given technology, and how to improve upon current hiring practices and procedures.

Program Outcome #12
The student will investigate, compare, contrast, analyze and form conclusions to current aviation, aerospace, and industry- related topics in unmanned aircraft and spacecraft systems, including UAS systems, robotics and control, unmanned systems operations and payloads, and human factors in unmanned systems.

To better understand UAS operations, this project will investigate requirements for UAS. This will include a detailed look at how UAS are classified, operational requirements, technology available, and future trends of UAS operations. A compare/contrast analysis will be completed for manned and unmanned aircraft in similar classifications to better understand why legislation for unmanned aircraft is stringent. Based off these findings a thorough analysis of UAS technology, mission needs, economics, and existing or anticipated challenges to routine use in NAS operations will be reviewed. Forecast effects of emerging technologies as well as anticipating new technological innovations in areas of airframes, powerplants, sensors, communication, command and control systems, and information technology and processing will be evaluated. Doing this analysis will provide better insight for future policy makers, technology manufacturers, and users of the NAS. With this review, the outcome will be to draw conclusions and recommendations on how to best use what UAS bring to the NAS, and how to assist in air traffic control workload going forward.

Introduction

Background of the Problem

The shortage in air traffic control personnel can be traced back to 1981 when President Ronald Regan fired more than 11,345 professional air traffic controllers (Macaray, 2015). The FAA’s hiring plan is designed to phase in new hires as needed over time. This will avoid creating another major spike in retirement eligibility in future years like the one resulting from the retirement of controllers hired after the 1981 controller strike. There are more than 14,000 air traffic controllers that work for the FAA, with an additional 1,297 contracted employees, and 9,990 military controllers. This workforce provides services within the United States to more than 24.6 million square miles of airspace (FAA (e), 2018).

With air traffic control, the workload an air traffic controller endures is dynamic; the dynamic nature of the job also affects staffing projections. A driving force behind air traffic control is workload demand. This demand is generated by commercial as well as non-commercial flights. UAS have also operated on a limited basis, but in recent years, their addition has made a significant impact on controllers projected workload, mainly from a complexity standpoint. Beginning in December of 2015, the FAA made registration for all UAS mandatory, because of this policy change, projections and forecasts for staffing adequate air traffic controllers has come up short. For the NAS to operate efficiently and safe, there needs to be sufficient qualified controllers to cover peak traffic which can be generated from weather, daily, weekly, or seasonal variations (FAA(e), 2018).

Air traffic controllers are restricted by hours of operation, requiring work schedules to accurately account for staffing for high and low-volume periods. This is unique, in the sense that it gives the FAA flexibility in staffing a facility dependent on volume and workload. Air traffic has hit a decline since the year 2000, but is expected to increase as UAS begin to be further incorporated into the NAS. The FAA relies heavily on traffic count data, which accounts for operations at a given location. This helps to forecast staffing. Because of the nature of small UAS (sUAS) the traffic count data that is used is misaligned.

UAS operations. As mentioned, UAS operations have only operated in the NAS on a limited basis. This limitation stems from regulatory concerns governing UAS operations. The primary focus of UAS operations has been on commercial grade UAS that operate in Class A airspace (airspace above 18,000 feet mean sea level). With UAS being integrated, an area that is often overlooked is low-level UAS (those that operate below 500 feet). This is problematic primarily because of the volume UAS bring. By the end of May 2017, there were nearly 772,000 registered owners of UAS that weigh between 0.55 pounds and less than 55 pounds also referred to as sUAS. By December of that same year there were almost 100,000 more registered users, with current projections putting that market to 3.17 million registered users in the United States by 2021. This influx of sUAS cannot reflect accurate data, because of reporting methodology. As mentioned earlier, the FAA utilizes traffic counts to predict for air traffic control staffing, but these operations can only be accounted for within controlled airspace (FAA, 2016).

For UAS operations, because of regulations, many sUAS do not operate in classes of airspace other than general, without prior written approval. This is where the problem of workload and volume become a factor with regards to UAS operations. Without technology to assist manage this workload, or account accurately for these sort of operations, the predictive staffing nature of “staffing to traffic” needs to be adjusted.

Economic impact of UAS. By 2025, it is estimated that UAS will be able to create nearly 100,000 new jobs to the U.S. economy. The economic impact of integrating UAS into the NAS will equivalate nearly $13.6 billion within the first three years, and looks to culminate nearly $82.1 billion by 2025. It is proposed that integrating UAS will create almost 34,000 jobs in manufacturing jobs and nearly 70,000 jobs related to UAS technology within the first three years as well. Most of the manufacturing jobs will be high paying ($40,000). Each year that UAS are not integrated, the United States loses out on nearly $10 billion in economic impacts. This number represents almost $27.6 million for every day UAS are not incorporated into the NAS (AUVSI, 2013).

Researcher’s Work Setting and Role

The researcher has a Bachelor’s Degree in Air Traffic Control from Thomas A. Edison State University (TESU) is currently working on his Masters of Aeronautical Science at Embry-Riddle Aeronautical University (ERAU). The researcher has 5 years of experience working as an air traffic controller at Naval Air Station Jacksonville. While serving in the United States Navy, the researcher was an air traffic manager, and fully qualified in the tower and radar environments. Through his experience, the researcher has faced problems when trying to integrate UAS into the NAS and has first-hand work experience as an air traffic controller.

Statement of the Problem

With air traffic control facing a 30-year low in fully qualified controllers, without measures in place, the NAS will not be able to increase capacity. Air traffic control workload is dynamic, and utilizing technology that is dated and does not offer automation, will affect future capacity in the NAS. On the horizon is the implementation of UAS, and this will have a profound effect on air traffic control operations and how the NAS is managed. The inability to fully staff facilities with fully certified controllers will thwart efforts to expedite UAS implementation. With a projection of nearly 3.1 million sUAS users over the next decade, the lack of updated technology, and the struggles of staffing facilities, what will occur without countermeasures in place is yet to be seen.

Significance of the Problem

Predictive models are needed to manage the risk of a high workload, but are generally difficult to develop. This difficulty stems from the notion of predicting workload is typically a multilevel problem. By 2021 the NAS is proposed to be fully updated by NextGen, by contrast, air traffic control staffing is at a 30-year low. Given the fact that there is expected to be almost 3.1 million registered users of sUAS by 2025, understanding air traffic control staffing now is paramount. The reason it is paramount is without proper staffing, regulatory concerns with the safety and efficiency of the NAS are held in limbo. The latency between staffing, updates to technology, and incorporating new air traffic control techniques complicates matters. Air traffic controllers rely heavily on standard operating procedures (SOP). Without the future generation of controllers being equipped to handle UAS operations, there will be a butterfly effect for years to come. Addressing staffing needs now before UAS become fully incorporated will help to adjudicate safety concerns and will help to model updates to the NextGen system.

Limitations

The first limitation this research brings is the limited data on sUAS operations. As the FAA begins allowing for these operations data will become more readily available. The second limitation is that predicting workload is subjective to the individuals’ tolerances. While the research can provide insight, the research itself is multi-faceted and lends itself to only being illustrative in nature. The third limitation is access to studies involving updated technology. With NextGen implementation, automation is expected to assist air traffic controllers and users of the NAS. While there are studies that suggest how technology and automation can impact a controllers’ workload, they do not cater to the technology being incorporated specific to NextGen and UAS operations. The final limitation is complexity of traffic at different airports. While research can illustrate an important human factor issue, it cannot illustrate all the variances and complexities seen at different airports.

Assumptions

This study assumes that data derived during simulation will replicate real world scenarios. Using data from human-in-the-loop simulations relies heavily on what the scenarios are built to analyze. In this research, the assumption is that data can simulate UAS operations and mimic their effect on an air traffic controllers’ workload. Other assumptions include the data derived will be enough to illustrate a human factor issues. It is not possible to account for every complexity in air traffic control, the independent variable in the research can be manipulated, but cannot account accurately for every complex issue encountered.

Review of Relevant Literature

In 2012 a Congressional Bill was introduced which was considered a milestone for UAS integration. With this Bill, deadlines and expectations for integrating UAS were laid out. This Bill also introduced waivers to initial airworthiness certification for UAS. This Bill is important to understand as it lays out the foundation for impacts to the NAS. Whether that be air traffic operations, flight standards, aircraft/aircrew certification, or even evaluation of current technology (Weibel & Hansman, 2005, p. 3). Understanding how this Bill impacts future operations, it is important to understand human factors related issues from both an air traffic control perspective, as well as a UAS operator perspective. Comparing these two items and collaboratively identifying areas that can be improved upon with technologies in place, human factors plays a vital role.

Within the framework of integrating UAS, it is critical to understand where UAS operate, how they are classified in their operation, and what technology could be utilized to mitigate concerns. On the other end of this spectrum is the concern of over automation. As will be mentioned in the human factors regarding air traffic control, automation may help workload, but is not viable solution, as it causes secondary human factors issues. The literature review will detail these issues and analyze some key talking points. To understand what can be affected, and how it will be affected, can bring light to solutions on how to mitigate these issues when integration inevitably occurs. With projections by 2025 of nearly four million registered users of “hobby-type” UAS, neglecting these concerns related to human factors, could cause a ripple effect in the NAS in the years to come.

ATC Human Factors

An area that is often reviewed when talking about operational performance is air traffic control. The personnel involved, how they handle workload, their interactions with their tools, these items are critical areas to understand when speaking about UAS integration. The significance of these problems varies from small to large, dependent upon the individual and what the individual is given to combat these human factors. To better understand what these elements are comprised of, the SHEL model will be utilized and explained in detail. Primarily SHEL represents Software, Hardware, Environment, and Liveware (Human component). Each component of the SHEL can be looked at as an interface the controller would operate on. The interaction between liveware and software is one interface, environment and software could be looked at as another interface, and throughout these varying interfaces there are matches and mismatches that make up the bigger element of human factors in ATC. Each match is just as important of any mismatch, because characteristics of each can be looked at to improve a human factor related issue, or could eventually lead to human error. Each interface has its very own definition and they are listed below (International Civil Aviation Organization, Human factors digest, 1993, pp. 34-40):

Software – This interface can be looked at as the standard operating procedures, i.e. the rules, written documents, and procedures controllers need to abide by. In this realm, software needs to adequately account for what controllers are expected to do, if software does not meet this demand, then it is considered ineffective software.
Hardware – For this interface, we would look at how the hardware was configured, the different control features, surfaces, what displays are available, and how they operate at a functional level i.e. does it tell the controller of an impending issue with one of the controls?
Environment - For this interface, we review areas such as the social and economic climate of an air traffic controllers working environment. Essentially the stimuli a controller is around throughout the day, are there impacts to the natural environment (ergonomics).
Liveware – One of the most important interfaces is Liveware as it focuses on the controller itself, how they interact with other controllers, pilots, maintenance personnel, this can be looked at as human to human interaction or the human element itself.

Human element (ATC). As mentioned above, there are interfaces in which an air traffic controller works within that define the SHEL model. The interface that has a direct correlation with how the other interfaces perform is the liveware interface. Within the liveware interface discussions about the human element are unveiled. To focus further on what that means it is important to focus on this liveware component and understand the complexities it brings to the other interfaces. Under this realm of liveware or the “human element” items to review include stress, boredom, error prevention, needs while at work, function of teams, attitudes, fatigue, confidence/complacency, and individual preferences.

Stress is an area that air traffic controllers are often exposed to, and need to endure. It is one of the most relatable elements of the SHEL interfaces. Stress is often caused by being overly saturated with tasks, the innate feeling of being timely, feelings of responsibility, and when their worksite does not provide adequate equipment or the equipment fails. Another area that comes up frequently in the liveware interface is boredom. While this area of the human interface has not been researched well, controllers thrive on tasks, it is when the tasks become overly automated that controllers begin making errors. To combat some of these errors controllers heavily rely on a job/task analysis which can predict errors. Many of these errors are generated from the devices controllers use, it is the machine and human interface that helps to predict a controller’s susceptibility to causing an error (International Civil Aviation Organization, Human factors digest, 1993, pp. 34-40).

Looking at the liveware to liveware connections, controllers rely on their team. Team functionality and needs at work play a critical role in day to day operations. When a team is operating at a high level, they do not need as much of the other SHEL interfaces. When a team is not functioning at a high level, in the instance of a trainee or equipment malfunction, the individual controller will require more of each SHEL interface. Another issue that plays into liveware is a controller’s attitude. Controllers live by an ethos, they are expected to be “perfect” because an error could cost lives. The team environment plays a factor in this attitude. A buy in of sorts, where controllers set a systematic attitude at the worksite. Each one of the SHEL interfaces contribute to this attitude, as well as the items mentioned already. When considering these elements, items like fatigue, confidence/complacency, and individual preference also become an operational factor (International Civil Aviation Organization, Human factors digest, 1993, pp. 34-40).

Each controller has individual needs at work, and thus each controller has their own unique ability and aptitude to complete a certain task, this plays a role in fatigue. In saying this, there is no one size fits all approach to workload management. Having said this, it means that a certain level automation for one controller, may not be necessary for another controller. This is important as we discuss integrating UAS and what tools become available with NextGen being implemented. Air traffic controllers tend to be great decision makers who continuously solve problems throughout a shift. Controllers depend on a certain level of confidence, whereby they are tested, but not so tested they cannot complete a task. On the alternative spectrum, if a controller is never tested, they become complacent. Lastly, individual preference determines that team environment/attitude at work. For automation to be effective, it would have to work for 95% of the team. There is no one size fits all approach when speaking of individual preferences, there must be enough flexibility in the automation process, to allow for these preferences to occur (International Civil Aviation Organization, Human factors digest, 1993, pp. 34-40).

Automation. Air traffic control technology, more specifically, automation technology, has been afforded the ability to advance steadily with a level of sophistication. A variety of techniques have been used to measure this automation. Within the complexities of air traffic control, automation has to account for data such as, data displays, failure detection and diagnosis, data/voice communication, trajectory profiles, and weather prediction. Having automation could supplement a controller in the decision making process, but never replace a controller. It is simply there as a tool to help a controller plan (Wickens, Mavor, & McGee, 1997).

While these technologies have advanced and become more sophisticated, they cannot replace the role an air traffic control plays when orchestrating the movement of aircraft. Each automation tool becomes crippled by the model used to develop them. Automation capabilities are only as good as what they are modeled after. The human element in this automation thus becomes an important piece of the puzzle. Controllers in this complex and ever changing environment have a need to act in a supervisory capacity over each system and subsystem. Controllers in turn, need to be able to step in when automation fails. What makes air traffic control a unique environment, and is favorable to human interaction, is how much more flexible, adaptable, and creative, we are as humans. A machine can assist the human, but cannot account for the unpredictable nature of air traffic control (Wickens, Mavor, & McGee, 1997).

Human Factors Considerations (UAS)

Systematically the air traffic control system as well as the NAS are sophisticated, but they are also very complex. When discussing integration of UAS, a key element is human factors, but just as important is understanding the degree of complexity. One of the principle concerns with UAS operations is establishing guidelines for how they will impact human factor related issues. UAS operations have many of the same facets to their manned counterpart, but there are also elements that are dissimilar. The dissimilarities are namely sensory isolation for the pilot operating the UAS. This isolation poses a variety of blockades that reduce human performance but also included within the realm of human factors issues for UAS operators are the following: loss of sensory cues; vital for flight controls, inability to scan the environment (the vehicle itself), and delays with communication and control of the UAS. There are some other outlying factors within the UAS community where an operator will operate multiple UAS, which lends itself to a heavy workload without the proper automation in place (McCarley, 2005).

Human factors issues. Within UAS operations there are varying degrees of automation, dependent upon platform and platform mission. UAS can operate almost identical to a manned aircraft (stick and rudder) for a manual set up, or UAS can be fully automated with very little user input other than acknowledgement. This varying degree of automation is helpful to understand because it is inverse to the automation needed for air traffic control. Just like in air traffic control automation lends itself to being beneficial, but can also be a cause in human error. Each level of automation contains an element of cognition. With the first level of automation, the manual mode, the UAS operator is under a heavy cognitive load. This level of automation is susceptible to factors of isolation, as well as communication delays, these all add to the cognitive workload of the operator. Conversely, examining the other end of the spectrum, full automation, it can be inferred that UAS operators lack the ability to intervene appropriately. The automation when a UAS is fully automated leads UAS operators in a degraded state of situational awareness and allows them to fall further “out-of-the-loop.” Operators of UAS need the same considerations when discussing automation, albeit there are different requirements, a lot of the same rules apply here (McCarley, 2005).

An integral part of integrating UAS is the ability to provide adequate separation from other aircraft. While this seems straightforward in practice, consider an environment where weather is degraded, consider when there is equipment related issues, or consider when task saturation displaces situational awareness. While in good weather and within line-of-sight (LOS), UAS operations can run smoothly, in general. Where this becomes a human factor related issue is when that sensory isolation kicks into overdrive. Primarily, when there is bad weather, or a UAS operator attempts to operate beyond-line-of-sight (BLOS), human factors considerations become even more paramount when consider the NAS. UAS operators rely heavily on their sensors, these sensors help operators detect, see, and avoid (DSA) other aircraft. As mentioned however, the ability to do so when weather is degraded becomes very difficult.

There are emerging technologies out to assist UAS operators with DSA, primarily through communication, navigation, and surveillance technologies supported by NextGen. ADS-B, which will be discussed later, plays an instrumental piece of DSA. While these elements help the case for integration, considerations still need to be made for non-cooperative anomalies (birds, aircraft not equipped with equipment, etc.). These functions become a requirement then vise an option. With these functions in mind, there are two very important human factors issues that arise from them (McCarley, 2005).

One expectation is that UAS operators are likely to operate in an error prone system. Saying that means automated target recognition can be imperfect, more specifically when the system is expected to generate early alerts (collision avoidance, provide an operator ample time and distance to evade). The second expectation of UAS operators is that they will be faced with a flawed DSA system. This is a big consideration because when DSA is not working effectively, the operator is under even more pressure to effectively monitor their other sensors. These concurrent responsibilities and expectations provide insight into how automation plays a factor and how to effectively manage DSA (McCarley, 2005).

Sensory isolation. Because operators of UAS are separated from the physical aircraft, sensory inputs are no longer received. In manned aircraft flight, sensory information is routinely fed to the pilot, UAS pilots do not receive this. This has a direct impact on a UAS operator’s environment, and they become dependent on information fed to them via sensors. This information comes from the datalink system that is coupled with the ground control station (GCS) that operators often work at. This datalink dependency could and often does, lead to degradation in imaging, a limited viewing range, and a lack of feel for aircraft characteristics. Without sensory cues such as sound, kinesthetic/vestibular information, and visual input, it is often difficult for UAS operators to have full situational awareness within their operating environment (McCarley, 2005).

Datalink delays. As mentioned before, datalinks play a crucial role in information fed to the UAS operator. One of the factors to consider is how profound the affect is on human factors when datalink delays occur. This delay in communication/feedback can seriously degrade a UAS operator’s performance. What makes this difficult is that there are no systems or methods to predict the varying degrees of delay. The inability to predict these delays leads a pilot to over compensate rather than give a fixed response based off aircraft characteristics, as we often see in manned aircraft. This delay could be the difference for controllers adequately separating aircraft or not. When controllers know characteristics of an aircraft and that aircraft’s performance, they have an expectation of what the aircraft will do. This anticipatory response that controllers use, can be drastically altered if a pilot receives a delay in feedback. These disruptions to the NAS make the system less predictable, meaning workload increases for all users of the system (McCarley, 2005).

Automation. With UAS there is a general expectation that some operators may have a need to control multiple UAS. This will require some form of automation from the UAS itself. The additional undertaking for a UAS operator will increase workload but decrease situational awareness (SA) unless automation is incorporated. This relates back to the human-machine interfaces. This principle relates back to the same ideas presented under air traffic control automation. What this implies that the ground control station (GCS) will require a friendly enough user interface that will not inhibit a UAS operator from operating multiple UAS at the same time (Fern & Shively, 2009).

Types of UAS operations

Understanding the domain in which UAS operate helps to identify the integration process as well as how they correlate to human factor related issues. The FAA outlined operations as either being civil operations, public operations, or model aircraft operations. UAS operators that fall under civil operations are governed by Title 49 of the United States Code, U.S.C. § 40102(a) (41) which provides the definition for “public aircraft.” § 40125 provides governance of appropriate qualifications to be appropriated with public aircraft status. Should this status not be granted, aircraft are then classified as civil operations. Under civil operations, operators may gain authorization in one of two ways. Either a Part 107 license or a special airworthiness certificate (SAC) (Huerta, 2013).

For pubic operations (governmental) the FAA authorizes operators a way to obtain a Certificate of Authorization (COA). When using a COA, the operator is granted permission to operate for their UAS for a specific time, place, and airframe. This airframe would have had to go through a screening to meet guidelines set out by the FAA, which would deem that airframe safe to operate in that airspace. Typical users that fall under this COA are entities such as law enforcement, border patrol, military training, and other sorts of Government related industry. Users are restricted to line-of-sight rules, and must not operate in a populated area ("Public Operations (Huerta, 2013).

Lastly, operations can be identified as hobby or recreation use. These sorts of operations fall under Subtitle B Section 336 of Public Law 112-95 formerly known as the FAA modernization and reform act of 2012. According to this doctrine, aircraft shall be flown strictly for hobby or recreational use, cannot weigh more than 55 lbs., shall not interfere with other aircraft operations, and will notify air traffic control if the aircraft will be flown within 5 miles of an airport. These operations will not require prior authorization if they are operating under these parameters (Huerta, 2013).

How are they doing it?

Unmanned aircraft have begun operations outside of the United States. This is an important piece to understand, because there is policy in place in those Countries that is allowing them to do this. While the United States is generally considered the busiest airspace in the world, aviation laws and regulations tend to play the same tune across the world. The main difference is the governing bodies. In the United States, Federal Aviation Regulations (FARs) are the rules to abide by, these are complemented by the FAA. So how do other Countries have the ability to operate unmanned aircraft? They pose the same challenges else ware, yet, Australia, Canada, Finland, Italy, Malaysia, Sweden, and the UK, have managed to create less restrictive regulations and appear to be miles ahead in integrating UAS into their airspace (Dalamagkidis, Valavanis, & Piegl, 2008).

ICAO (International Civil Aviation Organization) is an international aviation entity that has had involvement in integrating UAS into different countries since 2005. Their involvement in these various Countries helped to create several task forces, which inevitably led to those Countries developing regulations regarding UAS. Within those regulations, it was found that most of the Countries that have less restrictive regulations, review a “level of equivalent safety.” Using this method to evaluate UAS, Dalamagkidis, Valavanis, & Piegl (2008) stated the following:

“Regulatory airworthiness standards should be set to be no less demanding than those currently applied to comparable manned aircraft nor should they penalize UAS systems by requiring compliance with higher standards simply because technology permits.”

The United States seeks to do the same thing, but due to the vastly more complex nature of the U.S. NAS, it is difficult to apply the same principles. In addition to this, it is often hard to have an equivalent level of safety defined on UAS due to their uniqueness. An example with this would be regulatory concerns involving cockpits. Cockpits have a standard that is required by regulation to be met. In order to have an equivalent level of safety, there would need to be a standard for UAS regarding the cockpit.

This is where regulatory concerns and integration hit a plateau. To fully integrate UAS, a subset of regulations would need to be built upon specific to UAS, with their manned counterparts in mind. Operationally, they should have equivalent levels of safety, understandably, it is difficult to define this area of safety due to humans being detached from the aircraft. The primary difference between the United States and the other Countries mentioned, is the complexity of the airspace. The U.S. often sets a precedence on aviation regulations, which are then adopted by other countries, in this case, the United States. We may very well adopt some of the policies from countries successful in integrating UAS, but the United States will take a cautionary approach to their integration efforts (Dalamagkidis, Valavanis, & Piegl, 2008). Across all the Countries mentioned, their largest regulatory concern stems from incorporating small unmanned aircraft systems (sUAS).

sUAS concerns

Correlating these two areas of human factors we need to go back and review the Congressional Bill introduced in 2012. In this Bill an area that needs perhaps the most consideration is how to integrate small UAS (sUAS). In 2012, Congress allowed these operations to occur under an exemption, named Section 333. This has exemption has since become adopted and is known as Part 107 to Tile 14 of the Code of Federal Regulations (CFR). This addition to Title 14 of the CFR allows routine operations of sUAS and illustrates operating procedures and rules for those operators (Foxx, 2015). In these rules it is stated that pilots can fly one half hour before official sunrise, and one-half hour after sunset if the UAS is properly equipped with collision avoidance lights (red and green lights for identification). Additional requirements for these operations are that they: will take place in an area where visibility is at least 3 statute miles, the sUAS stays below 400 feet above ground level (AGL), they will not exceed a speed greater than 100 miles per hour (mph), and will not weigh more than 55 pounds (lbs.).

Risk associated with integration of small UAS. With these considerations, sUAS pose a different threat than larger UAS because of their operational characteristics. Larger UAS carry sensors that help detect, see, and avoid other aircraft, hence why they are not as limited. The size of sUAS plays a factor in DSA, even though their size is not as much a public concern, it is a concern when integrating them with the NAS. There are two concerns that raise risk with sUAS. The first one relates back to earlier mentions of human factors, and that is can the pilot effectively detect, see, and avoid other aircraft/objects? Can other pilots operating in the NAS detect, see, and avoid sUAS? The idea here is will the operator of a sUAS feel just as liable should something go wrong? The physical separation from the aircraft plays a factor in this decision (Foxx, 2015).

The other issue that creates risks echoes the same ideas of datalink delays from earlier. The systematic failure of any of the SHEL interfaces will lead to an interruption in the NAS. sUAS rely heavily on their datalink, and because of their size, are not nearly as robust as one would see on larger UAS. Lost link procedures vary across different sUAS, their programming or their “software” plays an important role as to what they will do. The inability to have effective control over this aircraft, and without safety measures in place is where workload on a controller becomes a concern. The inability to effectively predict a flight path, becomes a serious problem as we navigate how to effectively integrate not only UAS, but sUAS which will make up a majority of the market in years to come (Foxx, 2015).

Sense-and-avoid concerns. This cannot be reiterated enough, sense-and-avoid or see-and-detect, become the foundation for how sUAS will operate. As it sits now, there is not a technology available that will fully mitigate the paradigm of these issues. While we can create systems that will mimic what a pilot on board would see, what we lack still is sensory deprivation. Without providing sensory inputs to the pilot, and ensuring datalink delays are mitigated (to a point where they are almost negligible), sense-and-avoid concerns will be looming. The inability for sUAS operators to detect, see, and avoid aircraft like manned aircraft can do, will continue to set the precedence on their limitations. Per 14 CFR 91.113(b) states that “vigilance shall be maintained by each person operating an aircraft to see and avoid other aircraft.” This limitation, means that often sUAS require a visual observer and cannot operate BLOS (Foxx, 2015).

Lost-link concerns. Again, lost-link concerns represent a critical piece to integration. It is the loss of positive control of an aircraft that causes apprehensions for air traffic controllers and users of the NAS. The datalink is the sole source of an aircraft’s controls, without other safety measures in place, controllers and other users alike, will not be keen on allowing these operations to occur with the same fluidity as manned aircraft do. Manned aircraft that encounter emergencies, can often be mitigated by the pilot in charge (PIC) and controllers can identify these issues through communication, either visual or auditory (Foxx, 2015).

UAS Integration

For UAS, as with all aircraft, the FAA acts in a dual role. As the regulator, the FAA ensures aviation safety of persons and property in the air and on the ground. As the service provider, the FAA is responsible for providing safe and efficient air traffic control services in the NAS. Although aviation regulations have been developed generically for all aircraft, until recently these efforts were not done with UAS specifically in mind. This presents certain challenges because the underlying assumptions that existed during the previous efforts may not now fully accommodate UAS operations. The goal of safely integrating UAS without segregating, delaying, or diverting other aircraft and other users of the system presents significant challenges. The FAA recognizes that current UAS technologies were not developed to comply with existing airworthiness standards. Current civil airworthiness regulations may not consider many of the unique aspects of UAS operations. Materials properties, structural design standards, system reliability standards, and other minimum performance requirements for basic UAS design need to be evaluated against civil airworthiness standards for existing aircraft.

As UAS are introduced, their expected range of performance will need to be evaluated for impact on the NAS.

UAS operate with widely varying performance characteristics that do not necessarily align with manned aircraft performance. They vary in size, speed, and other flight capabilities. Similarly, the issue of performance gap between the pilot and the avionics will impact NAS operations. For example, a quantitative time standard for a pilot response to ATC directions (such as “turn left heading 270, maintain FL250”) does not exist – there is an acceptable delay for the pilot’s verbal response and physical action, but there is no documented required range of acceptable values. Avionics that perform the corresponding function cannot be designed and built without these performance requirements being established. Existing standards ensure safe operation by pilots actually on board the aircraft. These standards may not translate well to UAS designs where pilots are remotely located off the aircraft. Removing the pilot from the aircraft creates a series of performance considerations between manned and unmanned aircraft that need to be fully researched and understood to determine acceptability and potential impact on safe operations in the NAS.

Future Technology

An integral part of understanding how air traffic controllers will be affected by UAS operations will be the technology available. NexGen is perhaps the largest component in updating the NAS to modernize and advance aviation in the United States from a ground based surveillance system to a satellite based surveillance system. There are several components that encompass NexGen, but the one that has the most significant impact is Automatic Dependent Surveillance-Broadcast or ADS-B. This technology will be federally mandated to be equipped on almost all aircraft operating in controlled airspace with ADS-B out being required by 2020. Other technologies that are expected to help air traffic controllers and UAS operators are system wide information management (SWIM), National voice system (NVS), Terminal automation modernization and replacement (TAMR), Low-Altitude tracking and avoidance system (LATAS), Low Altitude Authorization and Notification Capability (LAANC), and Unmanned Traffic Management (UTM).

Air traffic control technology

Air traffic control is complex, understanding technologies that are being rolled out as part of the NextGen package, are important in helping to identify how they will increase productivity whilst decreasing workload. Part of this technology includes NVS, SWIM, and TAMR. Additionally, ADS-B plays an integral role of a larger picture when dealing with air traffic control. This is a high-level overview of what each technology will offer, identifying these items are crucial to help develop a case for controller and UAS operator interactions. The ability to understand these technologies will further reinforce how workload can be affected based off of these technologies. Perhaps the most important technology that will become accessible to controllers and UAS operators alike will be the use of ADS-B. These items will be addressed in the paragraphs to follow.

ADS-B technology will be the successor to many of the systems that are currently in place. This is a milestone technology that updates the NAS from older legacy RADAR systems to satellite based navigation. This technology will be used to provide data such as airspeed, location, and other data that is appropriate (aircraft identification etc.) This data would be received and broadcasted to a network of ground stations, in turn relaying that data to users such as air traffic controllers and UAS operators. Not only will it provide information about the status of a flight, but will also help to generate weather and traffic related information, deliverable to cooperating users. In saying cooperative users, it is those users that have the technology installed, it is a situational awareness tool. Two key elements of ADS-B help provide information critical to users of the NAS. They are ADS-B out, and TIS-B which will be explained in paragraphs to follow (FAA (a), 2018).

ADS-B out. One of the required components of ADS-B will be ADS-B out, required by 2020. ADS-B out is meant to replace legacy RADAR systems. In legacy RADAR systems, the sweep rate (rate which radar scans for aircraft) is every 5 to 12 seconds. With ADS-B out, this time is reduced to around a second. This increase in scan rate helps provide up to date information for all users. In the current system, radio waves can be obstructed, meaning long range communication is more difficult. ADS-B itself utilizes smaller stations, and will help to mitigate these issues. ADS-B provides a much better depiction of aircraft and their information, regardless of obstructions (FAA (a), 2018).

TIS-B. Traffic information service-broadcast (TIS-B) is like ADS-B. The optional component of ADS-B is ADS-B in which would allow other aircraft to “see” each other, so long as both aircraft are ADS-B equipped. In the event one aircraft is equipped and one is not, is where TIS-B comes into play. TIS-B uses an aircraft’s position and altitude from air traffic control radars, and converts that to a compatible format for ADS-B. This information is then broadcasted to other ADS-B users via one of two ADS-B datalinks (FAA (a), 2018).

Other important technologies for air traffic control include NVS, which will replace how voice switches operate at facilities. NVS allows for the FAA Telecommunications Infrastructure (FTI) to go from independent operations to shared operations. This technology will help to provide sustainable two-way communications, enhance monitoring capabilities, and provide a better means to route communications. It will allow air traffic control facilities to move communications to a neighboring facility should one facility be overworked due to weather. This will ensure continuous frequency usage (FAA, 2017).

Another critical piece of NexGen is SWIM. SWIM will allow air traffic management the ability to share information with other qualified systems. It is essentially the backbone of digital data-sharing in NexGen. SWIM Terminal Data Distribution System (STDDS) will take raw data from the surface data. This data will be put into information accessible by tower controllers to achieve a “big picture.” Essentially this is an intra-facility device that correlates what the Terminal Radar Approach Control (TRACON) is doing for sequence with the tower. Currently this system is available at 38 TRACONS, this system will team up with other components of SWIM to provide airlines and airports valuable arrival and departure information. This system also enhances flow control, which will balance traffic demands with capacity and allow controllers to better calculate end-to-end trajectories (FAA(c), 2018).

Lastly is TAMR. TAMR looks to replace outdated legacy RADAR systems within the air traffic control network. Figure 1. Illustrates the drastic differences between the two systems.

Figure 1. Comparison of ARTS to STARS (Old and New) RADAR systems. Retrieved from: https://www.faa.gov/air_traffic/technology/tamr/

TAMR is unique as it allows air traffic control systems to display digital “targets” versus analog targets. Targets in air traffic control represent aircraft, each “target” displays an aircraft with a primary and secondary “target.” Primary data is depicted as a blue rectangle, secondary data is the aircraft callsign, altitude, and indicated airspeed. Additional information can be annotated in the secondary data block by air traffic control input. STARS will also allow for six levels of precipitation to be displayed that range from light precipitation to extreme precipitation. This system allows for individual preferences to be set, as well as more accurate airspace depictions. Adaptations are performed monthly on STARS allowing for the display to be current. According to the FAA STARS is expected to be fully implemented by 2020 (FAA (d), 2018).

Air traffic management technology

Critical to the future infrastructure of the NAS is how air traffic will be managed, more specifically, how that airspace will be managed with UAS being incorporated. As more and more aircraft are put into the NAS, the ability to maintain performance at capacity will be dependent on future air traffic management technologies. Some of these technologies were mentioned in the previous section, ADS-B, SWIM, but this section will review systems like UTM, LATAS, and LAANC. As part of the technology overview, it is critical to understand what technology will be available to help manage UAS.

UTM. In May of 2018 NASA in conjunction with the FAA produced a proof of concept for UTM. Due to the large integration of low-altitude UAS operations, there are a variety of challenges that are presented. One of the solutions to these challenges is the use of UTM. This technology will be utilized predominately by sUAS. It will be an operator to operator based service. This means UAS operators will be responsible for coordination, management, and execution of their operations. This technology is expected to help assist air traffic controllers in both controlled and uncontrolled airspace by geofencing the operational area. In general, UTM is expected to service areas at or below 400 feet above ground level (AGL). While only a concept, this technology will seek to authorize and evaluate operations. As more data becomes available the system will be flexible and evolve to meet needs (Federal Aviation Administration, 2018).

LATAS. Like UTM, LATAS seeks to provide services to sUAS at low-altitudes as well. The primary difference is UTM will be a system whereas LATAS is more of a hardware package one could install on their UAS. This is a subsidiary program that is being utilized under the umbrella of UTM. The intent of LATAS is to provide better visual awareness to UAS operators at low-level altitudes. It will do so by providing a heads-up display (HUD) that depicts where other aircraft are, and provide heat maps for terrain and obstacles where UAS pilots are operating. The system itself is currently used in 128 countries and is being tested under the FAA’s Pathway program (LATAS, 2018).

LAANC. Another technology that is expected to be completely rolled out by September 2018 is the Low Altitude Authorization and Notification Capability (LAANC). This technology is a collaborative effort between the FAA and the UAS industry. It is meant to directly support UAS and their integration into the NAS. This technology will work to provide near real-time authorization requests for UAS operators to operate in controlled airspace. LAANC allows for the authorization process to be automated to receive approval. UAS operators would use this to apply for airspace authorization for operations under 400 feet in controlled airspace. Users that want to operate above this 400-foot restriction may apply up to 90 days early and receive manual authorization to do so. Working in a similar fashion to the other systems mentioned, this system further reduces workload by providing automation to authorizations. The ability for the system to do this automatically will assist controllers in the future (FAA (b), 2018).

Workload factors

To understand how these technologies will impact controllers, it is important to have the ability to estimate workload. Because of the complex domain of workload estimation, there have been several published studies. In past estimation studies, workload was focused in on air traffic controllers directly, the studies utilize traffic count and complexity matrices. The limitation to this is that these studies did not capture individual preferences. An alternative approach to evaluate workload is to focus on complexity factors. Researchers have worked on ATC workload measurement since the 1960s. The consensus is that ATC workload is affected by objective factors (e.g., traffic count, complexity of airspace/crossing routes) and human factors, such as those listed in the SHEL model. Amongst these factors, controllers work can be viewed as either visible work or invisible work (Song, Chen, Li, Zhang, & Bi, 2012).

The visible work would correspond to items such as observable work (recordable work). Invisible work would then represent a controller’s intellectual work, which is viewed as cognitive workload. Invisible work is harder to measure because it involves physiological and psychological factors. In air traffic control, if there are only a few aircraft, control actions are determined easily as potential conflicts are easily resolved. As the traffic count increases, along with complexity, workload increases due to cognitive workload (Song, Chen, Li, Zhang, & Bi, 2012).

Presently there are two methods that are effective at measuring workloads: DORATASK and MBB. In both methods control actions are classified first, with an evaluation placed at how quickly the control action takes place or the “instruction time.” These methods do have limitations (1) due to the changing environment of ATC, any evaluations that have been made, will have to be redone once change occurs, and (2) the methods do not factor in sector structure (complexity of airspace) (Song, Chen, Li, Zhang, & Bi, 2012).

Mogford et al. (1995) summarized 40 factors that affect complexity in the airspace. This study was built on previous research done in the field, and pointed out that these 40 factors are fundamentally the reasons for change that affect workload. Within the construct of their research they classified workload into five factor categories: static airspace, dynamic, human, equipment, and unexpected. Table 1 illustrates these categories.

Table 1

Workload Categories

Category	Factors
*Static airspace*	Number of airport terminals • Number of intersecting flight paths • Airline hub location • Special use airspace • Sector geometry • Sector size • Requirements for longitudinal and lateral spacing • Number of altitudes used
*Dynamic*	• Aircraft density: - Number of aircraft - Traffic distribution - Aircraft density or traffic volume • Conflict: - Presence of conflicts - Preventing conflicts (crossing or overtake) • Other： - Aircraft handled in prior time interval (e.g., last hour) - Number of arrivals - Number of emergencies - Number of special flights - Coordination - Traffic mix (arrivals, departures, and over flights) - Number of path changes - Traffic flow structure - Clustering of aircraft - Mixture of aircraft types - Climbing and descending aircraft - Number of military flights - Aircraft routing
*Human Factors*	• Staffing • Number of communications with aircraft • Number of communications with other sectors • Number of handoffs and printouts • Control adjustments involved in merging and spacing • Number of required procedures • Handling pilot requests
*Equipment Factors*	• Equipment Status • Radar Coverage • Frequency Congestion
*Unexpected Factors*	• Weather conditions • Weather and its severity • Other

Source: Mogford, R. H., Guttman J.A., Morrow S.L., et al. (1995). “The complexity construct in air traffic control: A review and synthesis of the literature.” DOT/ FAA/CT2TN95/22.

Summary

This literature review demonstrated several factors that affect air traffic controllers. An area of concern was illustrated at the beginning of this project, primarily the tremendous growth potential expected from UAS operations. This growth creates a demand for air traffic control services, and places a considerable amount of pressure on the NAS. The literature review illustrated what the task environment of an air traffic controller looks like and provides a better understanding of the necessity to monitor and control workload as traffic density increases. Characteristics of ATC complexity can be used to analyze how these factors can affect the controller’s workload.

Another critical element found during the literature review demonstrated how controller cognitive strategies, information displays related issues, and personal factors, can vary complexity factors across different personnel. Essentially stating that complexity is not experienced nor handled the same by each controller. This ties back into the SHEL model, whereby each controller has different interactions with each interface. These varying levels of complexity are dependent on sector structure (i.e. airspace complexity) as well as traffic flow patterns. In addition, the literature pointed to the overall quality of the human-machine interface. The human-machine interface quantifies how a controller will cognitively process and adopt strategies to complete a task based on that technology.

A variety of technologies were introduced when reviewing the literature, illustrating advancements that will be made in the future with NexGen. These technologies demonstrate a level of automation that is not currently in the NAS. As mentioned in the SHEL model, automation can be good if it is moderated. Too much automation will lead to complacency and will harm the NAS in safety, should issues arise. There were also several technologies that were introduced that would assist controllers with their workload when discussing UAS. Having technologies that can lessen the load for air traffic controllers in their dynamic environment will prove to be handy in the long run, as capacity increases and the workforce decreases.

Statement of Hypothesis

This statistical analysis will test if air traffic control will be affected by the incorporation of UAS. The hypothesis is that the incorporation of UAS will have an impact on controller performance.

Methodology

Research Model

A qualitative, inferential approach was used in this research study. The overall objective was to evaluate an air traffic controllers’ workload when UAS were incorporated with conventional (non-UAS) aircraft. Data for this was collected from a simulated study involving UAS. The construct of the model was to utilize a varying degree of UAS (1, 2, or 4) in combination with a variance in speed (slow, mixed, or fast), this was the independent variable in the research model. Each UAS was assigned a fixed delay response of 1.5 seconds. UAS could receive responses right away, but would have a delay in acknowledgement. The conventional aircraft in the simulation were pseudo-pilots and had a near immediate response time. In this research model the voice software that was utilized simulated transmissions where pilots could “step” on each other, meaning there was frequency congestion. Some transmissions would be received, some would have to be repeated for acknowledgement. Each of the simulations would run for 40 minutes.

This research model reflects the same criterion on workload measurement that was discussed in workload factors. The simulation itself involved static airspace (ZLA airspace) in addition to utilizing intersecting flight paths. Dynamic workload factors were also measured in the airspace by changing the speed and number of aircraft involved. This contributes to the density of traffic. Regarding the human factor element of the research model, measurements were made on hand-off times and controllers were expected to coordinate with each other. Equipment related issues involved a fixed communication delay with UAS, simulating the potential delays expected with future integration of UAS.

Research Population

The research population consisted of simulated data where air traffic controllers with previous experience were utilized (average of 25 years working air traffic control). Eight radar-certified controllers that were familiar with the airspace from previous work experiences, were selected to perform in the simulation. The simulation took place over two days.

Sources of Data

The data for this research was collected from a simulated study on air traffic control and UAS, performed by the National Aeronautics and Space Administration (NASA) Ames research center.

Reliability

The data utilized for this study is deemed reliable as it comes from a Government resource (NASA). This data also utilized certified professional controllers during the simulation with a sector that they had formerly worked in with a culmination of experience averaging 25 years.

Validity

Due to the shortage of available certified professional controllers and the impending boom of UAS being integrated with the NAS, this study is valid on the basis of better understanding the impact UAS have on controller workload. It is important to measure, even in a small group, as to provide better insight about performance and efficiency standards. While the study itself was over only a couple of days, the study itself does use controllers that are experienced and familiar with airspace. The validity of the data collected from this study is supported by comparing the study to human factor issues identified in previous research. This data creates a measurable variable on workload based on these factors.

Treatment of Data

The data for the study was retrieved from the simulation to make an analysis. The analysis involved using repeated measures ANOVA. ANOVA was used in lieu of t-tests due to type 1 errors that could occur utilizing a t-test.

Results

During the analysis, air traffic controllers were examined based on their performance for loss of separation events and the distance an aircraft traveled through their sector (safety vs efficiency). This is an important metric because it helps to identify what future technology solutions should be provided. This analysis is also critical in assisting to identify areas of human factor related issues that need to be mitigated or potential changes to air traffic management that need to be implemented because of the analysis.

Loss of separation averaged a mean of 2.6 with a standard error of 0.28. This number was not significantly impacted when UAS were present, but there was a trend that developed demonstrating that as the number of UAS increased, so did the occurrence of loss of separation. This is illustrated in figure 2.

Figure 2. Mean loss of separation per number of UAS.

This linear trend accounted for nearly 44% of the total variance in loss of separation, in addition to this there was significance in the loss of separation, as about half of the loss of separation events occurred with UAS involvement. Figure 2 presented F (1, 7) =5.42; with a p= .05.

Another observation made from this study was the significance in number of UAS x Speed interaction representing F (4, 28) = 5.026, p=.004 as represented in figure 3.

Figure 3. UAS x Speed interaction distance aircraft traveled through a control sector.

When one UAS was present in the sector, conventional (non-UAS) aircraft distance covered was not affected by UAS speed. When there were two UAS present, the distance that conventional aircraft travelled was higher if the UAS was at a slow speed in comparison to UAS at high speed or a mix of speeds. When there were four UAS present, UAS that travelled slowly would decrease the average distance travelled in comparing the four mixed and four fast UAS.

The last observation before controllers were evaluated for workload was the interaction of speed and number of UAS. This metric looked at average hand-off accept times and is demonstrated by figure 4.

Figure 4. Mean Handoff acceptance time in seconds for Number of UAS x Speed.

In figure 4, the main effect of workload interaction was modified by a significant number of UAS x Speed, with data representing F (4, 28) = 3.16, p=.029 respectively. Where there was only a single UAS in the sector, hand-off acceptance times were generally shorter so long as the UAS was slow. When there were two UAS present, at fast speeds, hand-off accept times were comparatively shorter than when there was two UAS present traveling at mixed speeds or slow speeds. When there were 4 UAS present, mixed and slow speeds trended towards hand-off accept times that were shorter comparatively to when there were constant fast speeds. This would indicate that workload tended to be lower for slow UAS if there were either one or four slow UAS present in the sector, but workload would increase if there were two slow UAS present. In this simulation, there is a strong indication that where there was a singular UAS in a sector, more aircraft would enter, which inevitably increased a “perceived workload.” Alternatively speaking, if there were four UAS present, less aircraft would enter a sector when comparing it to two UAS being present. When there were two UAS present, loss of separation occurred more frequently.

Conclusion

With research performed by NASA, and other industry leaders, the FAA is in collaboration in integrating UAS into the NAS with precautionary steps. Due to the explosive growth of UAS in the near future, and a lack of fully qualified controllers, the next five to ten years will need some serious updates to air traffic management principles. As detailed in this project, human factor related issues play a critical role, and knowing there is a potential for the NAS to be crippled to some extent, the need for research to continue in human factors related to UAS integration must be stated, and continually reviewed. The future of air traffic management needs to account for these human factor related issues. A risk based approach will need to be incorporated for a successful integration to occur.

NextGen offers several key technologies that will assist air traffic control personnel combat workload. With that being said, and perhaps more vital, is the concept of unmanned air traffic management. In the study that was conducted, controllers were already navigating manned aircraft around unmanned aircraft. If the unmanned aircraft could coordinate with themselves, and those operators, it would eliminate the controller in that loop. By doing this controllers can still plan and effectively work traffic, it is the delay and the level of automation that will be the pivotal turning point in future legislation regarding UAS.

There is a strong need for regulation specific to UAS. As detailed, the economy could greatly benefit from the use of UAS. The issue at bay will be addressing technology. There will be a need to have procedural improvements as well as a need to review and mitigate safety concerns. While integration should take place, it needs to take place at a pace that is consistent with staffing in the NAS. This in conjunction with technology updates should suffice until more research can be conducted.

Recommendation

Reviewing the items detailed in this project regarding human factors in air traffic control and integration, there are a few recommendations that would suit industry goals. As mentioned previously, regulation needs to be fair, regulation should not hinder progress. Albeit, UAS present different challenges, the NAS works very well for a reason. For integration to take place, and to assist in human factor issues, regulation must stay consistent, to a degree. While not all regulation will be consistent, due to the unique nature of UAS, there should be enough consistency that air traffic control can still work as intended.

Building on the concepts presented in this research project, a continual risk-based approach will be needed to have success. As more data becomes available, and further research can be conducted in to human factor related issues, more integration can occur. The industry i.e. UAS manufacturers, should take the lead in developing guidelines. These guidelines will need to be vetted, and demonstrated in simulations before making hasty regulation that would not allow for full integration. There is a continual need to simulate air traffic, this helps to synthesis data and create alternative solutions. Simulating data can be useful in creating better automation tools for controllers, it would need to be a meta-data analysis rather than small studies. To truly gain insight into areas that need to be addressed, there needs to be adequate data. Because new technology is on the horizon, and new traffic management solutions are being presented; it is highly recommended that each of these technological advances, are looked at with microscopic precision.

The scope of integrating UAS into the NAS is certainly a tall task. The NAS is not just adding a singular experimental aircraft such as the Condor. The NAS is seeking to add an entirely new category and class of aircraft. While the transiting for larger UAS will not be as noticeable, the addition of sUAS will be felt. Overall the driving force behind integration comes down to human factors. Can the NAS support integration effectively without overworking air traffic controllers? If it cannot, integration will need to be slowed until this can occur.

Future research recommendations would include review of NexGen with UAS. NexGen will play an instrumental role in the overall development of rules and regulations going forward. This project presented human factor issues, and utilization of the SHEL model should continue to stay present. The SHEL model in addition to what Mogford made observations of in air traffic control lays out the foundation for future research. As was mentioned, it is crucial to understand the impact of UAS, without data, there can be no progress. To truly capture these concepts, it is strongly recommended that while integration occurs, secondary systems exist that either take workload off controllers, or is supplemented by automation, but not too much automation.

References

AUVSI. (2013, March). The economic impact of unmanned aircraft systems integration into the United States (Rep.). Retrieved https://higherlogicdownload.s3.amazonaws.com/AUVSI/958c920a-7f9b-4ad2-9807-f9a4e95d1ef1/UploadedImages/New_Economic Report 2013 Full.pdf

Edwards, T., & Martin, L. (2017). The relationship between workload and performance in air traffic control. Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 61(1), 1609-1613. Doe: 10.1177/1541931213601887

FAA. “Unmanned Aircraft Systems.” Aerospace Forecasts, Federal Aviation Administration, July 2016, www.faa.gov/data_research/aviation/aerospace_forecasts/media/Unmanned_Aircraft_Systems.pdf.

Federal Aviation Administration. (2018). Concept of Operations v1.0 [Brochure]. Author. Retrieved from https://utm.arc.nasa.gov/docs/2018-UTM-ConOps-v1.0.pdf

FAA. (2017, April 03). National airspace system (NAS) Voice System (NVS). Retrieved from https://www.faa.gov/nextgen/programs/nvs/

FAA (a). (2018, January 08). Automatic dependent surveillance-broadcast (ADS-B). Retrieved from https://www.faa.gov/nextgen/programs/adsb/

FAA (b). (2018, June 20). FAA UAS data exchange. Retrieved from https://www.faa.gov/uas/programs_partnerships/uas_data_exchange/

FAA(c). (2018, February 06). System wide information management (SWIM). Retrieved from https://www.faa.gov/air_traffic/technology/swim/

FAA (d). (2018, April 09). Terminal automation modernization and replacement (TAMR). Retrieved from https://www.faa.gov/air_traffic/technology/tamr/

FAA (e). (2018). 2018-2027 Air traffic control workforce plan (United States, Department of Transportation, Federal Aviation Administration). Retrieved from https://www.faa.gov/air_traffic/publications/controller_staffing/media/CWP_2018.pdf

Fern, L., & Shively, R. J. (2009). A comparison of varying levels of automation on the supervisory control of multiple UASs. In Proceedings of AUVSI’s Unmanned Systems North America 2009, Washington, D.C.

Foxx, A., & Huerta, M. (2015, February 15). Operation and certification of small unmanned aircraft systems (Notice of proposed rule-making). Retrieved from https://www.faa.gov/regulations_policies/rulemaking/recently_published/media/2120-AJ60_NPRM_2-15-2015_joint_signature.pdf

Gilbert, T. (2016, June 15). Air traffic control staffing shortage must be addressed. Retrieved from http://thehill.com/blogs/congress-blog/technology/283328-air-traffic-control-staffing-shortage-must-be-addressed

Glaneuski, J. (2013, September). Unmanned aircraft system (UAS) service demand 2015-2035: Literature review and projections of future usage (United States, Federal Aviation Administration, Research and Innovative Technology Administration). Retrieved from https://fas.org/irp/program/collect/service.pdf

Huerta, M. (2013). UAS Operations in the NAS. In Integration of civil Unmanned Aircraft Systems (UAS) in the National Airspace System (NAS) roadmap (1st ed., pp. 14-21). Department of Transportation-Federal Aviation Administration.

Huerta, M. (2015, May 1). NextGen implementation plan 2015. Retrieved from https://www.faa.gov/nextgen/media/NextGen_Implementation_Plan-2015.pdf

International Civil Aviation Organization. (1993). Human factors digest (Vol. 8, Human factors in air traffic control). Montreal, Canada: International Civil Aviation Organization.

Dalamagkidis, K., Valavanis, K., & Piegl, L. (2008). On unmanned aircraft systems issues, challenges and operational restrictions preventing integration into the National Airspace System. Progress in Aerospace Sciences, 44(7-8), 503-519. doi:10.1016/j.paerosci.2008.08.001

LATAS. (2018). The drone safety platform. Retrieved from http://www.flylatas.com/

McCarley, J., & Wickens, C. (2005, April 1). Human factors implications of UAVs in the National Airspace. Retrieved from http://www.tc.faa.gov/logistics/grants/pdf/2004/04-G-032.pdf

Mogford, R. H., Guttman J.A., Morrow S.L., et al. (1995). “The complexity construct in air traffic control: A review and synthesis of the literature.” DOT/ FAA/CT2TN95/22.

Sondgeroth, J. R. (2018, June). FAA pathfinder project reveals challenges of shared airspace. Unmanned Systems, 36(4), 40-43.

Song, Z., Chen, Y., Li, Z., Zhang, D., & Bi, H. (2012). Measurement of controller workloads based on air traffic complexity factors. Cictp 2012. doi:10.1061/9780784412442.194

Weibel, R., & Hansman, R. (2005). An integrated approach to evaluating risk mitigation measures for UAV operational concepts in the NAS. Infotech@Aerospace, 3-10. Doi: AIAA-2005-6957

Wickens, C. D., Mavor, A. S., & McGee, J. P. (1997). Flight to the future: Human factors of air traffic control. Washington, D.C.: National Academy Press.

Scott Leishman's Unmanned blog

Sunday, July 2, 2023

SNHU BUS 400 Blog 1-2

Tuesday, October 29, 2019

Final Capstone Project (Unmanned Aircraft and ATC)