An Evaluation of Flight Safety Assessment and Management to avoid Loss of Control during Takeoff

Journal of Aerospace Information Systems

2015

Self Cite

This paper presents a novel flight safety assessment and management augmentation to the flight management system designed to assist a flight crew in avoiding or recovering from impending loss-of-control situations. Nominally, this system serves as a passive monitor but, in high-risk situations, warnings and (ultimately) override actions are initiated to mitigate the high-risk situation. In this work, flight safety assessment and management is applied to the task of preserving safety during takeoff, which is one of the highest-risk phases of flight. Flight safety assessment and management is specified as a deterministic Moore machine that can ultimately be certified using existing software certification processes. To facilitate understanding and to reduce state-space complexity, flight safety assessment and management's state machines are split into longitudinal and lateral-directional submachines that identify and mitigate loss-of-control contributing factors associated with aircraft dynamics and control constraints. Case studies based on documented takeoff accidents are presented to evaluate flight safety assessment and management's ability to maintain safe flight in realistic loss-of-control scenarios. Results from these case studies illustrate that flight safety assessment and management could have averted the takeoff accidents that were considered. A discussion of other factors that must be considered before realizing a comprehensive flight safety assessment and management capability for takeoff is provided. NomenclatureA lg = longitudinal takeoff logic A lt = lateral takeoff logic ap = envelope aware autopilot control C L , C D = lift and drag coefficients h = altitude p, q, r = angular rates p = pilot control T, W = thrust and weight u, v, w = velocities in the body frame V = true airspeed V lof = liftoff speed V R = takeoff rotation speed V 1 = takeoff decision speed X = longitudinal position on runway Y = lateral position on runway y = crosstrack error α, β, γ = angle of attack, sideslip angle, and flight-path angle δ a , δ r = aileron and rudder inputs ρ, μ, S ref = atmospheric density, friction coefficient, and planform area ϕ, θ, ψ = roll, pitch, and yaw angles

“…To enable a sensitivity analysis, we also chose different thresholds and crosswind magnitudes. In each scenario, FSAM consistently rejected the takeoff whenever possible [8].…”

Section: A Loss Of Directional Control In Continental Airlines Flighmentioning

confidence: 98%

Section: Takeoffmentioning

confidence: 99%

See 1 more Smart Citation

Flight Safety Assessment and Management for Takeoff Using Deterministic Moore Machines

Journal of Aerospace Information Systems

2015

Self Cite

“…LOC results when an aircraft exits its safe flight envelope or collides with another aircraft, building, or surrounding terrain and has been widely addressed in previous research [2][3][4][5][6][7][8]. In our previous publications [8][9][10], a flight safety assessment and management (FSAM) capability was proposed to prevent LOC during the takeoff phase of flight. FSAM is an automation aid responsible for real-time assessment of LOC risk, activation of risk-based warnings, and resilient control override of the flight crew if necessary.…”

Section: Nomenclaturementioning

confidence: 99%

“…FSAM is a high-level flight management system decision aid responsible for real-time assessment of LOC risk, activation of risk-based warnings, and resilient control override of the flight crew if necessary. In previous work, FSAM was applied to the prevention of LOC during takeoff [8][9][10] but the FSAM capability is still in early stages of development. Figure 2 illustrates a manually constructed DMM A: S; S 0 ; Σ; Λ; T ; G that represents FSAM logic for the longitudinal dynamics of takeoff.…”

Section: A Takeoff Flight Safety Assessment and Managementmentioning

confidence: 99%

Verification Guided Refinement of Flight Safety Assessment and Management System for Takeoff

Özay

Journal of Aerospace Information Systems

2016

Self Cite

Systems that make safety-critical decisions must undergo a rigorous verification and validation process to ensure automation decisions do not jeopardize the nominal safe state of operation. Flight safety assessment and management is a high-level decision-making system to reduce loss of control risk. This paper demonstrates how tools from formal verification can be used to guide the design of a takeoff flight safety assessment and management system implemented as a deterministic Moore machine. Finite state abstractions of simplified takeoff dynamics under different control authorities (i.e., pilot vs safety controller) are computed and composed with the Moore machine. By construction, the composition captures all behaviors of simplified takeoff dynamics. Then, a model checking tool analyzes whether this composition satisfies the takeoff safety requirements specified by federal aviation regulations. The results of model checking together with the abstraction are used to refine the Moore machine to ensure satisfaction of the specification. This paper contributes a novel approach to verification of a supervisory system specified by a Moore machine and applies this technique to flight safety assessment and management, which is itself an emerging flight management automation aid.

An Autonomous Override System to Prevent Airborne Loss of Control

2016

AAAI

Loss of Control (LOC) is the most common precursor to aircraft accidents. This paper presents a Flight Safety Assessment and Management (FSAM) decision system to reduce in-flight LOC risk. FSAM nominally serves as a monitor to detect conditions that pose LOC risk, automatically activating the appropriate control authority if necessary to prevent LOC and restore a safe operational state. This paper contributes an efficient Markov Decision Process (MDP) formulation for FSAM. The state features capture risk associated with aircraft dynamics, configuration, health, pilot behavior and weather. The reward function trades cost of inaction against the cost of overriding the current control authority. A sparse sampling algorithm obtains a near-optimal solution for the MDP online. This approach enables the FSAM MDP to incorporate dynamically changing flight envelope and environment constraints into decision-making. Case studies based on realworld aviation incidents are presented.