CONTEXT DEPENDENT TOTAL ENERGY ALERT FOR THE DETECTION OF LOW - - PowerPoint PPT Presentation

CONTEXT DEPENDENT TOTAL ENERGY ALERT FOR THE DETECTION OF LOW ENERGY APPROACHES

M A S T E R ’ S T H E S I S P R O P O S A L M I C H A E L P O R T M A N

• Commercial aviation operations have encountered a series of suboptimal approach

profiles, collectively named “unstable approaches”

• Consist of metrics including energy states
• Alerts exist which warn pilots of each individually
• However, no system exists which combines these metrics into a low total energy alert
• Such an alert is proposed in this thesis
• Integrates data already available onboard
• Estimates current energy state and trend in total energy
• Predicts whether the total energy will become too low, and alerts pilots with enough time for

corrective action

INTRODUCTION

• Much research has taken place evaluating the use of energy metrics in aviation
• No specific research on an alert for all modes of approaches
• Will determine the ability of alert to guard against low energy unstable approaches

better than current technology

• Application of FOQA data allows for larger scale analysis of energy alerts than previous

research

• Real world application and validation

CONTRIBUTIONS

Focus of literature review centers around five significant questions:

• 1. What is the function of an alert?
• 2. What is total energy?
• 3. What is context dependency?
• 4. Why would we need a total energy alert?
• 5. What are the attributes of a good total energy alert?

LITERATURE REVIEW

Pritchett (2001): “An alerting system is an electro-mechanical system capable of monitoring for, detecting and announcing conditions anticipated (by the operator or the system designer) to impact the operator’s near-term activities.”

• Properly detect certain conditions
• Announce the presence of these conditions
• Conditions are anticipated to impact the operator’s near-term activities
• Q1. WHAT IS THE FUNCTION OF AN ALERT?

Pritchett, 2001

• Kinetic energy (airspeed) + gravitational potential energy (height above ground)
• Energy naturally decays over the course of the approach (descent reduces potential

energy)

• Energy can be “transferred” between stores (by pitching)
• Energy is added by increasing thrust
• Low energy is when there is too little energy, system-wide, to respond by pitching
• Adding energy/thrust is a required response
• Q2. WHAT IS TOTAL ENERGY?

• Allows alert to be tailored to each individual circumstance
• Earlier alerting to allow more time to respond
• Better prediction
• Later desensitization/inactivation
• Consists of knowledge of:
• Phase of flight
• Modeled ideal aircraft approach profile (speed, altitude)
• Aircraft configuration
• Local conditions (e.g. terrain, wind)
• Q3. WHAT IS CONTEXT DEPENDENCY?

• Q4. WHY WOULD WE NEED A TOTAL ENERGY

ALERT? Asiana Flight 214 Unstabilized Approaches

https://www.sfgate.com/bayarea/article/72-passengers-reach-settlement-over-Asiana-crash-6113481.php http://www.gatco.org/gatco-news/2016/10/17/new-guidance-material-on-unstable-approaches-published/

• Crashed on approach into SFO
• Manual approach flown by pilots

used to Autoland

• Began with high approach
• Pilots tried to lower approach,

resulting in low energy unstable approach ASIANA 214 OVERVIEW

Boeing, 2014

• NTSB investigation cited lack of pilot awareness of autoflight functions and lack of

situational awareness

• Recommendation to FAA (A-14-43): Task a panel of human factors, aviation operations,

and aircraft design specialists…to develop design requirements for context-dependent low energy alerting systems for airplanes engaged in commercial operations. ASIANA 214 INVESTIGATION

NTSB, 2013

FSF ALAR Stabilized Approach Definitions:

• Aircraft is on the correct flight path
• Only small changes in heading/pitch required to maintain correct flight path
• Aircraft speed is close to Vref (No more than Vref + 20 kts, no less than Vref)
• Aircraft is in the correct landing configuration
• Sink rate <1000 feet per minute
• Power setting appropriate for set configuration
• Briefings and checklists conducted
• Special circumstances accounted for and briefed

UNSTABILIZED APPROACHES AND STABILIZED APPROACH CRITERIA

FSF, 2000

• Alert:
• Direct pilot’s attention to the predicted low total energy state of the aircraft’s approach
• Total Energy:
• Measure stability based on evaluation of both altitude and airspeed metrics together
• Context Dependent:
• Be aware of phase of flight, aircraft configuration, and local conditions to individualize alerting

threshold

• Good:
• Better able to identify low total energy approaches than current onboard systems
• Allows more time for pilot corrective action
• Additional Considerations
• Q5. WHAT ARE THE ATTRIBUTES OF A GOOD

TOTAL ENERGY ALERT?

• Currently installed and operational
• Autothrottle Wakeup
• Autothrottle feature which advances throttles when airspeed is sensed to be too low
• Inactivates in certain autopilot modes
• EGPWS
• Monitors aircraft location with respect to terrain, glide slope
• Desensitized below 150ft radar altitude on approach
• LAA
• Alerts when 30% into amber band
• Inactivated below 200ft radar altitude on approach

REVIEW OF CURRENT TECHNOLOGY

Boeing, 2014

Several broad categories of research surrounding energy metrics:

• General Aviation education applications (PEGASAS)
• Commercial Aviation post-flight analysis
• Energy based automatic flight control systems
• Energy based alerting for autoflight approaches

REVIEW OF CURRENT LITERATURE

Puranik, et al., 2016; de Boer, et al., 2014; Lambregts, 1985; Shish, et al., 2015; Shish, et al., 2016; FAA, 2013; FAA, 2017

Design Overview:

• Designed explicitly for use during approach and landing phases of flight
• Monitors kinetic energy (airspeed) and above ground gravitational potential energy

(radar altitude), as well as their decay over time

• If decay is predicted to cause too little energy in the amount of time needed to safely

abort the approach, an alert sounds PROPOSED SYSTEM

PROPOSED SYSTEM Mathematics:

• 1. 𝑈𝐹=𝑛𝑕​𝑨↓𝑢𝑓𝑠 +​1/2 𝑛​𝑊↑2
• 2. ​𝑒𝑈𝐹/𝑒𝑢 =𝑛(𝑕​𝑨 ↓𝑢𝑓𝑠 +𝑊𝑏) or ​𝑛𝑕​𝑨

↓𝑢𝑓𝑠 +𝐺𝑊

1. F = T-D

• 3. ​𝑢↓𝑡𝑏𝑔𝑓 =​𝑢↓𝑡𝑞𝑝𝑝𝑚 +​𝑢↓𝑠𝑓𝑏𝑑𝑢𝑗𝑝𝑜↑∗

1. ​𝑢↓𝑡𝑞𝑝𝑝𝑚 = −4.55​ln(1−(​𝐸/𝜀𝑈 ))

• 4. ​𝑈𝐹↓𝑛𝑗𝑜 =𝑛𝑕​𝑨↓𝑠𝑓𝑟↑∗ +​1/2 𝑛​

𝑊↓𝑛𝑗𝑜↑2∗

1. ​𝑨↓𝑠𝑓𝑟↑∗ =𝑒 ∗𝑢𝑏𝑜𝐻𝑇

• 5. ∆𝑈𝐹≈(​𝑒𝑈𝐹/𝑒𝑢 )∗​𝑢↓𝑡𝑏𝑔𝑓

1. ​𝑈𝐹↓𝑔𝑗𝑜𝑏𝑚 =𝑈𝐹+ ∆𝑈𝐹

• 6. If ​𝑈𝐹↓𝑔𝑗𝑜𝑏𝑚 ≤ ​𝑈𝐹↓𝑛𝑗𝑜 , the alarm

sounds

• 7. If Tcommanded is sufficiently high such

that the system predicts the aircraft will recover, the alert will be silenced*.

• Functions for both manual and automated approaches
• Context dependency allows for proper response (including thrust) from pilots
• Alerts earlier than current onboard systems, giving more time to respond
• Allows for later ability to alert than systems without dynamic thresholds (deactivates

later) PROPOSED BENEFITS

ALERTING CONSIDERATIONS

• Sensor metrics
• Sensor lag
• Frequency of data capture
• Built-in uncertainty
• Aircraft performance
• Engine spool time
• Time to arrest descent
• Human factors
• Reaction time
• Dependent on many factors
• Primary reaction
• Immediate go-around reaction

trainable, reducing needed time

• Decision making allowance?
• Allows pilots to evaluate for themselves

the condition of the aircraft

• Necessitates earlier warning
• Previous pilot knowledge of event
• Potential for nuisance alarm
• However, justifies go-around call

• Start with application of alerting criteria to Asiana 214
• Preliminary application completed
• Validate preliminary research
• Application to large set of FOQA data
• Data collected from major American carrier
• Approaches into SFO and LAX
• Similar approach profiles and aircraft to Asiana 214
• Determine which flights trigger alert, compare to:
• Current FOQA unstable approach flagging
• Current onboard alerting systems
• Deeper analysis into flights which trigger alert

PROPOSED WORK

• Application/modification of real world influencing factors:
• Human reaction time
• Variation of total energy parameters
• Minimum allowable airspeed/altitude
• Thrust commanded vs. Thrust actual
• Time permitting:
• Evaluation of non-standard approaches
• Phased alerting
• More complex configuration effects in contextual awareness
• Determination of optimal sensor suite characteristics for alert
• Sensor limitations

PROPOSED WORK

• Application of alert to Asiana 214
• Graphical analysis shows alert

would have sounded ~22-27 seconds before impact

• 0.8-1nmi before runway
• Approximately doubles advanced

warning time WORK TO DATE

Boeing, 2014

Abbott, K., McKenney, D., & Railsback, P. (2013). Operational use of flight path management systems—Final report of the Performance-based operations Aviation Rulemaking Committee (PARC)/Commercial Aviation Safety Team Flight Deck Automation Working Group (CAST). Commercial Aviation Safety Team Flight Deck Automation Working Group.

• Boeing. (2014). Boeing submission for Asiana Airlines (AAR) 777-200ER HL7742 landing accident at San Francisco - 6 July 2013 (M. E. Bernson, Comp.).

de Boer, R. J., Coumou, T., Hunink, A., & van Bennekom, T. (2014). The automatic identification of unstable approaches from flight data. In 6th International Conference on Research in Air Transportation, ICRAT (pp. 26-30). Federal Aviation Administration (FAA) (2013). Safety Alert for Operators: Manual Flight Operations. Retrieved from: https://www.faa.gov/other_visit/aviation_industry/airline_operators/airline_safety/safo/all_safos/ media/2013/SAFO13002.pdf Federal Aviation Administration (FAA) (2017). Safety Alert for Operators: Manual Flight Operations Proficiency. Retrieved from: https://www.faa.gov/other_visit/aviation_industry/airline_operators/airline_safety/safo/ all_safos/media/2017/SAFO17007.pdf Flight Safety Foundation. (2000). FSF ALAR briefing note 7.1—Stabilized approach. Approach and landing accident reduction toolkit. Hansman, R. J., Wanke, C., Kuchar, J., Mykityshyn, M., Hahn, E., & Midkiff, A. (1993). Hazard alerting and situational awareness in advanced air transport cockpits. Haslbeck, A., & Hoermann, H. J. (2016). Flying the needles: flight deck automation erodes fine-motor flying skills among airline pilots. Human factors, 58(4), 533-545. Lambregts, A. A. (1985). U.S. Patent No. 4,536,843. Washington, DC: U.S. Patent and Trademark Office. McFadden, K. L., & Towell, E. R. (1999). Aviation human factors: a framework for the new millennium. Journal of Air Transport Management, 5(4), 177-184. National Transportation Safety Board. (2013). Descent below visual glidepath and impact with seawall: Asiana Airlines Flight 214. Pritchett, A. R. (2001). Reviewing the role of cockpit alerting systems. Human Factors and Aerospace Safety, 1(1). Puranik, T. G., Harrison, E., Min, S., Jimenez, H., & Mavris, D. N. (2016). Energy-Based Metrics for General Aviation Flight Data Record Analysis. In 16th AIAA Aviation Technology, Integration, and Operations Conference (p. 3915). Shish, K. H., Kaneshige, J., Acosta, D. M., Schuet, S., Lombaerts, T., Martin, L., & Madavan, A. N. (2015). Trajectory Prediction and Alerting for Aircraft Mode and Energy State Awareness. In AIAA Infotech@ Aerospace (p. 1113). Shish, K., Kaneshige, J., Acosta, D., Schuet, S., Lombaerts, T., Martin, L., & Madavan, A. N. (2016). Aircraft mode and energy-state prediction, assessment, and alerting. Journal of Guidance, Control, and Dynamics, 40(4), 804-816.

REFERENCES

Questions? Special thanks to Dr. Pritchett, Dr. Feron, and Dr. German for their input, support, and feedback. THANK YOU FOR YOUR TIME