Scholarly article on topic 'Situation Awareness for Mid-air Detect-and-avoid System for Remotely Piloted Aircraft'

Situation Awareness for Mid-air Detect-and-avoid System for Remotely Piloted Aircraft Academic research paper on "Earth and related environmental sciences"

CC BY-NC-ND
0
0
Share paper
Academic journal
Procedia Manufacturing
Keywords
{"Situation awareness" / ATC / RPA / "Collision avoidance" / Detect-and-avoid}

Abstract of research paper on Earth and related environmental sciences, author of scientific article — Jens Alfredson, Petra Hagström, Bengt-Göran Sundqvist

Abstract As part of the European Detect and Avoid project MIDCAS three human-in-the-loop simulation campaigns were conducted with remotely piloted aircraft system equipped with a Detect and Avoid system flying according to instrument flight rules in an air traffic control environment. Situation awareness for remotely piloted aircraft system pilots and air traffic controllers was studied as part of an evaluation of the operational concept. Experienced pilots and air traffic controllers participated. Data was collected for qualitative analysis including observer logs, questionnaires, and notes from debriefings/workshops. In general the situation awareness was good however there were also examples of lost situation awareness during the simulations, reported in this paper. Lessons learned include how the system could further support remotely piloted aircraft system pilots situation awareness by highlighting if intruder is transponder equipped or not and semi-static or frozen information regarding traffic avoidance and standardized phraseology should be used to further support shared situation awareness between remotely piloted aircraft system pilot and air traffic controller. Given that remotely piloted aircraft systems are new in this kind of setting the result from this simulations are promising in that there were no major concerns found concerning situation awareness that would risk significantly affect the operational concept.

Academic research paper on topic "Situation Awareness for Mid-air Detect-and-avoid System for Remotely Piloted Aircraft"

CrossMark

Available online at www.sciencedirect.com

ScienceDirect

Procedia Manufacturing 3 (2015) 1014 - 1021

6th International Conference on Applied Human Factors and Ergonomics (AHFE 2015) and the

Affiliated Conferences, AHFE 2015

Situation awareness for mid-air detect-and-avoid system for

remotely piloted aircraft

Jens Alfredson, Petra Hagström, Bengt-Göran Sundqvist

Saab Aeronautics, Bröderna Ugglas gata, SE-58247Linköping, Sweden

Abstract

As part of the European Detect and Avoid project MIDCAS three human-in-the-loop simulation campaigns were conducted with remotely piloted aircraft system equipped with a Detect and Avoid system flying according to instrument flight rules in an air traffic control environment. Situation awareness for remotely piloted aircraft system pilots and air traffic controllers was studied as part of an evaluation of the operational concept. Experienced pilots and air traffic controllers participated. Data was collected for qualitative analysis including observer logs, questionnaires, and notes from debriefings/workshops. In general the situation awareness was good however there were also examples of lost situation awareness during the simulations, reported in this paper. Lessons learned include how the system could further support remotely piloted aircraft system pilots situation awareness by highlighting if intruder is transponder equipped or not and semi-static or frozen information regarding traffic avoidance and standardized phraseology should be used to further support shared situation awareness between remotely piloted aircraft system pilot and air traffic controller. Given that remotely piloted aircraft systems are new in this kind of setting the result from this simulations are promising in that there were no major concerns found concerning situation awareness that would risk significantly affect the operational concept.

© 2015 Published byElsevierB.V. Thisisanopen access article under the CC BY-NC-ND license

(http://creativecommons.Org/licenses/by-nc-nd/4.0/).

Peer-review under responsibility of AHFE Conference

Keywords/Situation awareness; ATC; RPA; Collision avoidance; Detect-and-avoid

1. Introduction

As part of the European Detect and Avoid (D&A) project MIDCAS (Mid-air Collision Avoidance System) [1,2], three human-in-the-loop simulation campaigns have been successfully performed with a remotely piloted aircraft system (RPAS) equipped with a D&A systemflying according to instrument flight rules (IFR) in an air traffic control (ATC) environment.

2351-9789 © 2015 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license

(http://creativecommons.Org/licenses/by-nc-nd/4.0/).

Peer-review under responsibility of AHFE Conference

doi: 10.1016/j.promfg.2015.07.161

The overall purpose of the project was to identify adequate technology, contribute to standardization and demonstrate a D&A system for RPAS able to fulfill the requirements for self-separation (or traffic avoidance) and mid-air collision avoidance in non-segregated airspace. Evaluations are performed using Real Time, Monte Carlo and ATC simulations as well as flight tests with a RPAS equipped with a D&A system. The main purpose of the ATC simulations was to study the D&A system from an operational perspective and the interaction with ATC. As part of this purpose situation awareness (SA) for remote pilots and ATCOs was investigated. Specifically, we here report on how active air traffic controllers (ATCO) and remote pilots were studied systematically in a real time simulation. Hereby we can report results on actual and desired behavior of the studied concept, regarding SA. The simulations where performed in three simulation campaigns carried out in Sweden during one week each. The focus of the simulations performed was on evaluating interaction between the D&A system, remote pilot, ATC and surrounding Air traffic during mid-air collision incidents. The subsystems included in the simulation were 1) Sense subsystem 2) Avoid subsystem (traffic avoidance and collision avoidance), see Fig. 1, and 3) Human Machine Interfaces. The performed simulations also included studies of workload [3] as well as assessments of the human machine interaction, not reported here.

In Fig. 1 below a notional timeline for the concept under study is presented for a typical situation which builds up. Depending on the chain of events and depending on simulated pre-conditions such as airspace class variations in expected behavior among the participants in the simulation would be expected.In the traffic avoidance phase the remote pilot can chose to activate the suggested maneuver, whilst if no action is taken by the remote pilot eventually, in the collision avoidance phase, the system will initiate an automatic maneuver. At any time of the situation buildup the conflict can be resolved by for instance intruder pilot action, ATCO instruction.

To provide an example of a typical chain of events the high level description below could help to understand one possible outcome. However there are several other chains of events that would be expected depending on for instance airspace class or if there are aircraft present without transponders.

1. D&A system: Provides information to remote pilot in order to allow him/her to establish SA. Checks RPA separation from other traffics (based on D&A separation minima fulfilment).

2. Remote pilot: Establish his/her own SA.

3. ATC: Under surveillance responsibility, ATC detects potential conflict with an intruder, which can be either IFR or VFR traffic. The characteristics of the intruder (aircraft type, speed, altitude, heading), its state (eventual contingency or emergency) and flight plan / intentions may also be known.

4. ATC informs remote pilot of potential conflict and provides him/her with an instruction to change the RPA heading and/or speed and/or flight level in order to maintain the separation from intruder according to appropriate rule. The intruder traffic's pilot is simultaneously informed via radio.

5. Remote pilot: Remote pilot acknowledges the instruction given by ATC via radio, and executes these instructions. Remote pilot monitors the correct efficiency of the de-conflicting action execution thanks to the support from D&A system situational display.

6. ATC: ATC informs the intruder Pilot of potential conflict and provides him/her with instructions; e.g. to maintain or to modify heading and/or speed and/or flight level.

7. Intruder pilot: Intruder pilot acknowledges the instruction given by ATC via radio, and executes instructions.

8. D&A system: Acting continuously as a complementary system, the D&A system continues to check that D&A separation minima is maintained. If a separation conflict is foreseen by the D&A system, it provides the remote pilot with a warning indicating the potential conflict and a resolution solution.

9. Remote Pilot: remote pilot informs ATC on the detected conflict.

10. ATC: ATC checks the remote pilot warning in order to verify conflict information. If the conflict is confirmed, then a new instruction will be given to the remote and intruder Pilots. If no conflict is found by ATC, they will inform the remote pilot to proceed as previously indicated. Note: If there would be aircraft present without transponders there are a significant risk that ATC may not be aware of them at all times.

11. Remote pilot: Remote pilot proceeds with the given instruction and informs ATC on the compliance with such instruction.

Fig.1. A timeline for the concept under study, for a typical situation. First, there is a surveillance phase and then a traffic warning or traffic alert is presented to the remote pilot. Then, there are traffic avoidance including nominal traffic avoidance and an escape phase. Finally after a collision avoidance alert the collision avoidance phase is entered. Then an automatic collision avoidance maneuver is initiated and the closest point of approach is passed.

An example of how the D&A graphical user interface could look during the simulations can be seen in Fig. 2, below. The main differences between the traffic avoidance and collision avoidance capabilities are that the Traffic avoidance has the objective for not scaring others whereas CA is designed not to scrape paint. Traffic avoidance is performed using nominal maneuvering performance (predictable for ATCO) whereas CA is using most of the RPA performance. Traffic avoidance is done by a change in one of altitude, heading or speed, whereas collision avoidance is done by a combined optimum maneuver, potentially in both horizontal and vertical direction. Traffic avoidance needs to be according to clearance in controlled airspace whereas collision avoidance is regarded as an emergency maneuver implying that ATCO only is informed after the initiation of the maneuver and after "clear-of-Conflict". Traffic avoidance needs to be commanded by pilot whereas collision avoidance is automatically triggered unled aborted by the pilot. This implies that collision avoidance is always available, independent of availability of data link, whereas traffic avoidance capability is lost when the link is lost.

Fig.2. An example of what the remote pilot could see at the D&A system HMI.

1.1. Situation awareness

Recently a holistic framework was proposed for SA [4]. However, the concept of SA has been debated for more than 20 years now [5,6,7] and are still under debate [8,9].During this time a huge amount of studies has used the concept within various domains, also for air traffic management. Of special interest to this study is that good SA is important for the ATCOs ability to handle unexpected aircraft behavior [10].

The SA of the ATCO is of interest for these simulations since the effect on the ATCO having to control RPAS equipped with D&A systems is of relevance for potential future technology, training and regulations to support safe and efficient performance.

The SA of the remote pilot is of interest for these simulations since it provided useful experiences of an operational context that could not be studied in real life at this point in time. Also it is interesting due to lessons learned of value for future requirement design, for instance of a D&A system for future RPAS.

Also shared SA between the ATCO and the remote pilot is of interest as a factor for explaining potentially unsafe behavior.

2. Method

The main method used in this study was human in the loop simulations. The simulations included valid scenarios performed by ATC experts and tested during pre-tests (Dry-Run).

Data was collected for qualitative analysis including: Observer logs by observers for ATCO and remote pilot, questionnaires, and notes from debriefings/workshops. Debriefings were held after each run performed and notes were collected by observers. Assessment of SA was specifically performed on remote pilots and air traffic controllers. The SA was both assessed by subjective assessments by the remote pilot and the traffic controller

respectively, but also though questions in a questionnaire and in debriefings and workshops performed together with subject matter experts.

The three ATCOs participating in the simulations were all active ATCO in the real airspace that was simulated in the study, one with nine years of experience in that position, another with two years' experience in that position, and the third had 27 years' experience in that position. The fourremotepilots were all experienced IFR pilots. One was a flight instructor and Saab 2000 pilot, another other had 33 years of experience as military and civil pilot, the third had about 2000 flight hours, and the fourth an active and experienced SAAB 340 and SAAB 2000 pilot with good familiarity with the simulated airspace.

Each team consisted of one ATCO, one remote pilot and two intruder pilots (controlling all manned aircraft and responding to ATCO communication) along with an additional "Rest of the world"-position, controlling the CTRs and the other adjacent airspace around the TMA, in order to make the exercise realistic with regard to handover between controllers. Five to seven one hour runs were performed with each team. There was also an initial reference run for each team with only manned aircraft, carried out for comparison.

Contextual information was carefully provided to the participants since potential misunderstandings of clearances might depend on certain flights or certain waypoints [11]. The simulated airspace under ATCO responsibility was Ostgota Terminal Maneuvering Area (TMA) in the south of Sweden, see Fig. 3. There were three active airports/CTRs under the TMA, the Linkoping/ Saab CTR, the Norrkoping/Kungsangen CTR, and the Stockholm/Skavsta CTR. Airspace classes C, D and E were simulated, whereas the true class of the TMA is C. Traffic density was also increased compared to current situation to enable assessment of a potentially more dense airspace in the future or on other parts of Europe.

The main difference between the simulated airspace classes with regard to Mid-Air collision prevention are the ATCO awareness and responsibility, which results in different responsibility of the pilots of the different aircraft. In airspace C the ATCO provides separation for all aircraft flying according to IFR against all other intruders. In this airspace class the remote pilot (IFR) is only responsible for providing collision avoidance. In airspace class D the ATCO is in contact and thus aware of all aircraft but is only responsible for providing separation between different IFR aircraft. The remote pilot will thus be required to provide traffic avoidance against VFR aircraft (and CA against all) under the condition that all aircraft are known to ATCO. In airspace E only IFR aircraft have to be known to ATCO and separation provision is thus only provided between these. The remote pilot will similarly be required to perform traffic avoidance against VFR aircraft (and CA against all) under the condition that all aircraft are not known to ATCO. Traffic information is also provided by ATCO in the airspaces where separation is not provided.

Fig.3. The simulated airspace was Ostgota TMA in the south of Sweden.

3. Results

The subjective SA ratings were consistent with data collected in the questionnaires, as well as the outcome of the debriefings and workshops.

In general the SA was good. At a direct question to the remote pilot "Does the information from the D&A system sufficiently support remote pilot's situation awareness?", the immediate response was "Yes, good".

During a situation that did not include any conflict and could be classified as normal the remote pilot was found by the observers to be very active with the graphical user interface of the D&A system zooming in and zooming out. This kind of behavior was common during these simulations. This particular time occurred on the third simulation week in the late afternoon the first day of simulation for this remote pilot. During the simulation various scenarios were used by varying the airspace class, varying if it would generate traffic avoidance or collision avoidance, if aircraft without transponders should be present and the flight plan of RPAS and surrounding traffic. The scenario was designed to study RPAS self-separation through the D&A system's traffic avoidance. This scenario explored the situation in which RPAS was expected to operate in an airspace class where ATC is not responsible for separation. RPAS was on an IFR flightplan, operate through D airspace on a clearance and would at various stages come into conflicts with VFR intruders, all of which are to be resolved by D&A system's traffic avoidance function well before collision avoidance function would activate. The duration of the scenario was 60 minutes but after 28 minutes of simulation this particular situation occurred, right after resolving a potential traffic avoidance situation for the second time during this run, but now being in-between events. In the debriefing the remote pilot explained the active zooming behavior as essential for him to build SA and he explained how he strived to have excellent SA also in situations when he did not expect any potential conflicts to occur. He stated that "It is good to be prepared".

Fig.4. A Zoomed out view at the D&A system's HMI of a situation in which the remote pilot did not have full SA.

However, there were also examples of lost SA during the simulations. At a specific situation the remote pilot was not aware of the situation at hand. There was a potential conflict. The RPAS observer noted this to bring it up during the debriefing after the run could clarify that the remote pilot had lost SA. The D&A system then performed an automatic collision avoidance maneuver according to its specified behavior and no collision was recorded. This occurred on the third simulation week in the early afternoon the first day of simulation for this remote pilot. It was the same remote pilot as in the situation described above and the adjacent simulation run. The scenario was designed to study RPAS D&A system's collision avoidance. This scenario was almost identical to the scenario described earlier in its pre-conditions, also airspace class D, similar flight plan, similar traffic density although some variations in traffic. Unlike the other scenario this scenario was designed to include the unpredictable behavior among transponder equipped as well as non-transponder equipped aircraft. The duration of the scenario was 60 minutes and after 40 minutes the situation of interest occurred. The situation is depicted in Fig. 4, below, with a zoomed out view. There is one aircraft right in front of the RPA and going in the same direction, and another aircraft in also in front but further away going in the opposite direction. The first aircraft to catch the remote pilot's attention was the aircraft right in front of him. He zoomed in with the intention to get better SA of that situation, but never gained SA of the other aircraft. As coming to know during the debriefing the incoming aircraft was the highest threat.

To support shared SA between the remote pilot and ATCO we found that two important criteria for the D&A system is to present:

1. Semi-static or "frozen" information regarding traffic avoidance manoeuvre direction, however needed to be reactive if scenario changes and initial manoeuvre suggestion is no longer safe. This enables the remote pilot to get new clearance from ATCO since non-emergency manoeuvring can only be done according to clearance in controlled airspace

2. Standardized phraseology and information content specification regarding traffic avoidance in line with current semantics, however too simple manoeuvres may result in large deviations from flight plan/mission.

To support remote pilot SA the participants of the simulation agreed in a workshop that the technical system should highlight to the remote pilot if a potential intruder is equipped with a transponder or not. This was motivated by supporting the SA by the remote pilot as well as supporting the SA by the ATCO through communication with the remote pilot.

4. Discussion and conclusion

SA was mostly good during these simulations both regarding the remote pilot and the ATCO. We learned how the system could further support remote pilot SA by highlighting if intruder is transponder equipped or not. This enables the pilot to understand if the ATCO, using transponder interrogation/Secondary Surveillance Radar, is likely to be aware of the intruder.

We also learned how to further support shared SAbetween remote pilot and ATCO by the use of semi-static or frozen information regarding traffic avoidance as well as the need for standardized phraseology.

Two situations were presented in the results section in which the remote pilot in both cases zoomed with the intention of increasing his own SA. In one of the situations he stated that this helped him in building SA but in the other situation it is likely that the zooming may have contributed to the lack of SA. It is up to further research to find out how to best support the need for zooming but reduce the risk of having negative effects of this on SA.

In encounters with non-transponding intruders in airspace C and D (triggered by failed transponder or erroneously strayed intruder) the ATCO queried the remote pilot for bearing and flight level of traffic to enhance the SA and give traffic information to surrounding traffic of the potential hazardous situation.

Given that RPAS are new in this kind of setting and that there are state-of-the-art methods to enhance safety within the domain of air traffic control [11,12] when they will be in use and we all can learn from practice, the result from this simulations are promising in that there were no major concerns found concerning SA that would risk significantly affect the operational concept.

References

[1] J. Pellebergs, The MIDCAS project. In Proceedings of the 27th International Congress of the Aeronautical Sciences, ICAS, Stockholm, 2010,

pp. 3241-3247.

[2] J. Pellebergs, MIDCAcS: The European detect and avoid project. In Book of Abstracts of the 4th international congress of the European

Aerospace Societies, Linkoping University Electronic Press, Linkoping Sweden, 2013, p. 189.

[3] J. Alfredson, P. Hagstrom, B.-G. Sundqvist, J. Pellebergs, Workload of a collision avoidance system for remotely piloted aircraft, In

Proceedings of the 46th Annual Nordic Ergonomics Society Conference, IEA Press, pp. 819-824.

[4] J. Lundberg, Situation awareness systems, states and processes: A holistic framework, Theoretical Issues in Ergonomics Science 02 (2015)

[5] M.R. Endsley, Situation awareness global assessment technique (SAGAT), In Proceedings of the IEEE National Aerospace and Electronics

Conference, 1988, pp. 789-795.

[6] J. M. Flach, Situation awareness: Proceed with caution, Human Factors 37 (1995) 149-157.

[7] M.R. Endsley, Towards a theory of situation awareness in dynamic systems, Human Factors 37 (1995) 32-64.

[8] M.R. Endsley, Situation awareness misconceptions and misunderstandings, Journal of Cognitive Engineering and Decision Making 9 (2015)

[9] J. M. Flach, Situation awareness: Context matters! A commentary on Endsley, Journal of Cognitive Engineering and Decision Making 9

(2015) 59-74.

[10] M.R. Endsley, M. Smolensky, Situation awareness in air traffic control: The picture, In E. Stein (Ed.) Human Factors in Air Traffic Control, Academic Press, New York, 1998, pp. 115-154.

[11] S. Ternov, G. Tegenrot, R. Akselsson, Operator-centred local error management in air traffic control, Safety Science 42 (2004) 907-920.

[12] S. Ternov, Learning for safety in health care and air traffic control. Doctolral dissertation 1650-9773, Lund University, Lund Sweden, 2011.