Rough Draft Peer Review

Human Factors in Unmanned Aerospace Systems

Embry-Riddle Aeronautical University

1 December 2019

Abstract

In most recent years, the rapid advancement of aviation technology has led to the establishment of a new type of aerial platform known as Unmanned Aerospace Systems (UAS), however, even though many various sectors such as military, agriculture, law enforcement, aerial photography have benefited tremendously from their usage and advantages, many major human factors issues continue to exist to this day. The relationship between human-machine interface has long been a controversial topic specifically in military UAS environments. Here, researchers demonstrated the effects caused by poorly designed MQ-9 Ground Control Stations (GCS) and the impact to pilots and sensor operators. Fifty consenting Air Force MQ-9 pilots and sensor operators from Creech Air Force Base, NV were asked to participate in a study involving GCS design limitations. Overall, the study concluded that poor cockpit design, fatigue, limitations to see-and-avoid capability, and lack of auditory cues were among the most significant human factors that affected performance. Additionally, further research is required in the area of GCS design improvement in order to provide a more user-friendly interface between operators and the UAS itself.

Keywords: Unmanned Aircraft Systems, human factors, MQ-9, GCS

Summary

The history of unmanned aviation can be tracked all the way back to the early 1900s, when Orville Wright and Charles Kettering were placed in charge of supervising a secret project that helped develop the world’s first self-flying aerial torpedo - deemed the “Kettering Bug” - which could fly autonomously to a pre-determined target and detonate. As the years passed by and technology advanced, so did Unmanned Aircraft Systems (UAS). Just like civilian aviation has come a long way since the Wright Brothers’ first flight in 1903, so has military aviation, and the same can be said about unmanned aviation. More specifically, the US military has relied heavily upon UAS to conduct Intelligence, Surveillance, and Reconnaissance (ISR) missions around the globe so much, that most operations nowadays are 24/7. In the Air Force, one such platform is the MQ-9 Reaper. Built by General Atomics and first flown in 2001, the MQ-9 is considered a medium-altitude Remote Piloted Aircraft (RPA) with a turboprop engine, 66-foot wingspan, and loiter time of over 25 hours. Capable of carrying a mixed load of armament weighing up to 3,800 pounds, the MQ-9 is considered an extremely reliable aircraft that in many cases meets or even exceeds manned aircraft reliability standards. The crew requirement for operating an MQ-9 Reaper is two - one pilot that is responsible for flying the aircraft and one Sensor Operator (SO) that operates all sensors and payloads on board. To do this, the crew operates from a Ground Control Station (GCS), which can either be inside a fixed facility such as a typical room in a building, or in a separate, mobile structure that resembles a container - some of which can be configured to house two crews inside that can operate two separate aircraft at any given time. The GCS, however, comes with its own set of limitations, most of which can be attributed to human factors.

Issue

In a study conducted on 25 November, 2019 at Creech Air Force Base, Nevada, 50 pilots and SOs from the same squadron were tasked to list the human factors challenges they faced in the GCS that negatively interfered with the performance of their duties. Of note, the sample size taken had an average experience level of 1,000 flight hours and 50% of the pilots polled had previous experience in flying manned aircraft; this specific statistic helped provide an effective comparison between manned and unmanned human factors issues. Overall, the study concluded that MQ-9 crews are experiencing multiple human factors issues in the GCS, ranging from design limitations and lack of sensory cues that need to be addressed and mitigated properly in order to ensure the high productivity of the crew and the success of the mission.

Significance of Issue

The study revealed that MQ-9 crews are experiencing multiple human factors issues that range from poor GCS design, fatigue, limited visibility, and lack of auditory cues. Crews stated that the design of the Pilot/SO (PSO) workstations is not user-friendly, as the two major complaints were poorly positioned aircraft controls and auxiliary display monitors too far apart from the central heads-up-display (HUD). Furthermore, the condition lever on the PSO workstation is in a position that could easily cause an engine shutdown if the operator is not careful with proper hand placement. The study also revealed that the lack of the seat-of-the-pants feel makes it harder for crews to analyze a situation such as turbulence, compared to manned flying. While manned pilots can more easily detect a stall or unusual attitude, unmanned pilots have to rely solely on electronic information, as there are no mushy feelings of the controls, buffeting, or kinesthesia present in the GCS that will warn them.

The Air Force currently uses two different types of GCSs for the MQ-9: the Block 15 and the Block 30. The Block 15 is the older version and is not as advanced as the Block 30, however, this paper focuses on the Block 15 since many units have not yet transitioned to the newer version and the differences between the two are minute. A typical MQ-9 GCS is configured with two identical PSO racks with the pilot in the left seat and the SO in the right. As pictured in Figure 1, the center screen on each rack is the aircraft HUD which displays telemetry such as airspeed, altitude and engine gauges. It is through this screen that the pilot primarily operates the aircraft - it is the central hub for the instrument crosscheck. Above the HUD is another screen called the tracker display that shows the aircraft on a constantly updated moving map. Additional settings can be found on this screen such as link status and aircraft radio frequencies. Directly below the HUD are two smaller touch-screens called the Heads-Down-Displays (HDDs) that show various aircraft system parameters such as engine, electrical, fuel, and weapons data. The crew can select different displays by typing in a corresponding number to that display. For instance, if the pilot wishes to see the status of all fuel tanks and the quantity in each one, then they would type 48 (48 is the fuel status menu) and all parameters would show up on the HDD. Two additional screens on the left side of the pilot rack and right side of the SO rack respectively, display information such as maps with other active aircraft in the airspace, secure internet chat rooms that enable communication through written means, and aircraft checklists. Finally, two screens in between the two racks display information such as landing times, fuel consumption rates, and other mission planning tools. A telephone is also located in between the racks, as well as the aircraft and ground radios. As seen, there is a lot of information readily available to the crew, however, this increases the amount of time spent crosschecking all the data which can be distracting at times.

Figure 1. General Atomics Legacy Ground Control Station. Adapted from “Legacy GCS,” 2014, General Atomics Aeronautical Systems. Retrieved from http://www.ga-asi.com/legacy-gcs

Each workstation encompasses a keyboard, mouse, control stick (joystick) on the right side and four levers on the left side. Both workstations are identical design-wise in case the pilot side rack malfunctions and has to use the SO side. The functions of each workstation’s controls are different as the pilot’s side controls the aircraft while the SO side controls payload operation and settings such as iris, camera type, and camera zoom. Specifically, the condition lever, which is located between the flap and throttle lever on the pilot’s side, controls the engine and allows fuel flow to the engine when in the forward position, closes or stops fuel flow (shuts the engine down) in the middle position, and feathers the propeller blades to reduce drag in the aft position (Carrigan, 2015). Having that said, out of all the levers and switches in the GCS, the condition lever is the most critical one since it directly affects engine operation. The research, however, revealed that the location of the condition lever is in a poorly chosen area since pilots can accidentally bump the lever with their arm when trying to manipulate the auxiliary screen on the left-hand side of the PSO rack. One pilot even revealed that his sleeve got caught on the lever as which moved it full aft, however, he was quick to react and immediately placed it back in the forward position before the command link could reach the aircraft. Furthermore, the study revealed that the condition lever can be easily mistaken for the flap lever due to its close proximity and same color. In certain cases, such as an in-flight emergency that requires engine shutdown either due to an engine fire or failure, the checklist will direct crews to place the condition lever in the aft position in order to feather the propeller blades and reduce drag. This provides the pilot with a better glide ratio and preserves as much altitude as possible. However, due to the close proximity and same color of the flap lever, the pilot can easily mistake it with the condition lever which can be detrimental during an emergency where time is of the essence. If the condition lever is not pulled in time, the propeller will not feather and the high drag that is created will severely impact the glide ratio. In other words, the aircraft will quickly descend out of the sky, losing much-needed altitude. Interestingly enough, only the speed lever knob - which is located to the right of the throttle and controls engine revolutions-per-minute - is painted red; all the other levers are painted black. This can further enhance the confusion between the flap and condition lever. As seen in Figure 2, the location of the condition lever in close proximity to the flap lever and the similarity in color make it easier for crews to confuse the two which, in emergency situations can prove to be costly.

Figure 2. PPO Setup with Condition Lever. Adapted from Human Factors Analysis of Predator B Crash. Retrieved from https://hal.pratt.duke.edu/sites/hal.pratt.duke.edu/files/u13/Human%20Factors%20Analysis%20of%20Predator%20B%20Crash%20.pdf

Another human factor issue is the relatively colder temperature inside the GCS compared to outside ambient temperature. Due to the various amount of equipment such as communication boxes, electrical panels, and displays that require constant, adequate cooling, the indoor temperature of a GCS is lower than what most humans are comfortable with; in some cases, around 64 - 67 degrees. Vimalanathan and Babu (2014) concluded that indoor room temperature has a 38 percent effect on performance, health, and productivity of office workers (Vimalanathan & Babu, 2014). Furthermore, the equipment also produces constant noise that creates distractions and makes it difficult to listen to the radio. Research has also shown that daytime noise exposure had a sustained effect on nighttime sleep, including shorter deep sleep and lower sleep efficiency (Guo, et al., 2017). Because of this, crews noted that the GCS causes fatigue which negatively affects performance. Due to the large amount of displays and long periods of endurance in the seat, research revealed that eye strain, shoulder and lower back pain, and headaches were some of the side effects. Because the displays are not located closely to the main HUD, crews have to constantly move their eyes back and forth as part of their normal crosscheck. According to human factors engineers, the three zones (i.e. “cones”) of visual location are “Easy Eye Movement” (foveal movement), “Maximum Eye Movement” (peripheral vision with saccades), and “Head Movement” (Kamine, 2008). In a study conducted by NASA’s Dryden Flight Research Center, Kamine & Haber (2018) measured instrument display visual angles to determine how well conventional aircraft and the MQ-9 ground control station (GCS) complied with these standards, and how they compared with each other (Kamine & Haber, 2018). It was discovered that all conventional vertical and horizontal visual angles lay within the cone of “Easy Eye Movement” and some in the “Maximum Eye Movement”, however, most instrument vertical visual angles of the MQ-9 GCS lay outside the cone of “Easy Eye Movement” (Kamine & Haber, 2018). In other words, the majority of MQ-9 GCS visual displays lay outside the cone of “Easy Eye Movement” which can cause eye strain. Reduced blinking rate and symptoms of eyestrain in operators of Visual Display Terminals (VDT) is not something new. Yakaishi and Namada (1999) concluded that reduced blinking rate, eyestrain, and uncomfortable eyes are more prevalent among VDT operators compared with office workers doing comparative jobs not involving VDTs (Yakaishi & Namada, 1999).

Compared to a manned aircraft where the pilot is physically located in the seat, UAS operators are located hundreds or even thousands of miles away from the aircraft and lack several sensory cues such as ambient visual input, kinesthetic, vestibular, and auditory information (Damilano, et al., 2012). The limited field of view, image resolution, and refresh rate - constrained by the data-link bandwidth - make it difficult for a UAS operator to see-and-avoid other aircraft in the sky. Additionally, since there is no seat-of-the-pants feel, crews indicated that the lack of sensory cues limits their ability to detect turbulence or erratic engine operation/vibration. As opposed to a manned pilot that can easily sense turbulence, erratic engine operation, unusual attitudes, or vibrations, UAS operators must rely on other sources of information - mainly electronic; the sense of balance and equilibrium provided by the inner ear is absent.

Recommendations

It is evident that many human factors challenges exist in MQ-9 Block 15 GCSs, however, there are many recommendations that could be implemented in order to improve crew performance. One recommendation is that the condition lever should be placed in a position that will allow quicker identification and separation from the other levers. Given the importance of this lever, it should be placed in an isolated location on the workstation, away from other controls. Also, installing a guard switch over it will ensure that it is not inadvertently pulled back by the pilot’s arm or sleeve. Additionally, color-coding the condition lever such as bright yellow and black will ensure that in times of emergencies, less time is spent trying to identify the lever and more time is spent trying to handle the emergency.

A second recommendation is that auxiliary display monitors should be placed as close to the HUD as possible. This will allow for an easier and quicker crosscheck for the crews, as well as alleviate any eye strain caused by excessive eye movement. The HUD itself should also be modified to provide a wider field of view width that will increase situational awareness. By displaying a wider horizon, crews can more easily detect and avoid other aircraft in the sky and weather phenomena such as potential cloud formations, lightning, and thunderstorms.

A third recommendation would be to include warnings of critical anomalous events that

involve more than one type of sensory mode such as both an auditory and visual warning of critical anomalous events (Williams, 2008). For example, in addition to a visual indication of engine RPM, providing the pilot with the option of listening to the actual engine noise would tremendously assist with detecting any unusual sounds. That option could be as simple as clicking a button and instantly listening to the aircraft engine whenever the pilot choses to do so.

Lastly, similar to manned aircraft, installing a stick shaker can help aid the pilot in recognizing an impending stall.

Conclusion

The MQ-9 GCS has come a long way since its original inception by incorporating various changes to its design, however, research has shown that there are still many human factors challenges that crews are facing. Therefore, additional research on human factor implications on MQ-9 GCSs must be conducted that will allow for continuous updating and refinement of cockpit design, monitor placement, audio-sensory cueing, and human-machine interaction. This would require additional funding, however, it is an absolute necessity if crews are expected to perform at their highest. As technology continues to improve, GCSs must be constantly refined, while always taking human factors considerations into account, in order to keep up with the constant demands placed on UAS operators.

References

Bendrick & Kamine (2019). Instrument Display Visual Angles for Conventional Aircraft and the MQ-9 GCS. Retrieved from https://ntrs.nasa.gov/search.jsp?R=20080022357

Damilano, Guglieri, Quagliotti, & Sale (2012). FMS for unmanned aerial systems: HMI issues and new interface solutions. Journal of Intelligent & Robotic Systems, 65(1-4), 27-42. doi:http://dx.doi.org.ezproxy.libproxy.db.erau.edu/10.1007/s10846-011-9567-3

General Atomics Aeronautical Systems Inc. (2019). Predator B RPA. Retrieved from http://www.ga-asi.com/predator-b

Guo, Lin, Tsai, Lin, Chen, Chung, & Wu (2017). 0429 Daytime workplace noise exposures lower than occupational criteria can disturb nighttime sleep. Occupational and Environmental Medicine, 74 doi:http://dx.doi.org.ezproxy.libproxy.db.erau.edu/10.1136/oemed-2017-104636.354

Haber, J., & Chung, J. (2016). Assessment of UAV operator workload in a reconfigurable multi- touch ground control station environment. Journal of Unmanned Vehicle Systems, 4(3), 203+. Retrieved from https://link-gale- com.ezproxy.libproxy.db.erau.edu/apps/doc/A463514960/AONE?u=embry&sid=AONE &xid=1d00ac6c

Nakaishi, H., & Yamada, Y. (1999). Abnormal tear dynamics and symptoms of eyestrain in operators of visual display terminals. Occupational and Environmental Medicine, 56(1), 6. doi:http://dx.doi.org.ezproxy.libproxy.db.erau.edu/10.1136/oem.56.1.6

Perez, D., Maza, I., Caballero, F., Scarlatti, D., Casado, E., & Ollero, A. (2013). A ground control station for a multi-UAV surveillance system: Design and validation in field experiments. Journal of Intelligent & Robotic Systems, 69(1-4), 119-130. doi:http://dx.doi.org.ezproxy.libproxy.db.erau.edu/10.1007/s10846-012-9759-5

Vimalanathan, K., & Babu, T. R. (2014). The effect of indoor office environment on the work performance, health and well-being of office workers. Journal of Environmental Health Science & Engineering, 12, 1-8. Retrieved from http://ezproxy.libproxy.db.erau.edu/login?url=https://search-proquest- com.ezproxy.libproxy.db.erau.edu/docview/1559854911?accountid=27203

Williams, K. (2008). Documentation of Sensory Information in the Operation of Unmanned Aircraft Systems. Retrieved from https://libraryonline.erau.edu/online-full-text/faa- aviation-medicine-reports/AM08-23.pdf