Flight safety operations optimization
Techniques for optimizing flight safety operations for an aircraft are described. In operation, aircraft operational parameters corresponding to a plurality of flight operations are retrieved. The aircraft operational parameters are then analyzed using a first machine learning model to identify a first flight operation, where the first flight operation comprises at least one aircraft operational parameter with deviation beyond a threshold. At least one potential flight safety incident corresponding to the first flight operation is then identified using the at least one aircraft operational parameter. The at least one potential flight safety incident is then analyzed using a second machine learning model to identify a corrective action for the potential flight safety incident, where the second machine learning model is trained using flight safety artifacts comprising a plurality of flight safety incidents and corrective actions to be initiated in response to the plurality of flight safety incidents. The corrective action is then subjected to an avionics digital twin to ascertain that the corrective action mitigates the at least one potential flight safety incident. The corrective action is then recommended for the at least one potential flight safety incident.
1 . A method for optimizing flight safety operations for an aircraft, comprising:
retrieving aircraft operational parameters corresponding to a plurality of flight operations;
analyzing the aircraft operational parameters using a first machine learning model to identify a first flight operation from amongst the plurality of flight operations, wherein the first flight operation comprises at least one aircraft operational parameter with deviation beyond a threshold;
identifying at least one potential flight safety incident corresponding to the first flight operation using the at least one aircraft operational parameter;
analyzing the at least one potential flight safety incident using a second machine learning model to identify a corrective action for the potential flight safety incident, wherein the second machine learning model is trained using flight safety artifacts comprising a plurality of flight safety incidents and corrective actions to be initiated in response to the plurality of flight safety incidents;
subjecting the corrective action to an avionics digital twin to ascertain that the corrective action mitigates the at least one potential flight safety incident; wherein subjecting the corrective action to the avionics digital twin comprises:
simulating the first flight operation with the corrective action on the avionics digital twin;
recording aircraft operational parameters corresponding to the first flight operation; and
ascertaining that the corrective action mitigates the at least one potential flight safety incident upon determining the deviation in the at least one aircraft operational parameter to be within the threshold;
recommending the corrective action for the at least one potential flight safety incident;
computing flight operation characteristics associated with simulation of the first flight operation with the corrective action, wherein the flight operation characteristics comprise flight time, fuel consumption, and Air Traffic Control (ATC) approval confidence score corresponding to the flight operation, and the ATC approval confidence score is indicative of chances of approval of the flight operation by the ATC; and
rendering the corrective action along with the flight operation characteristics.
2 . The method of claim 1 , wherein the aircraft operational parameters are retrieved from a Quick Access Recorder (QAR) database storing flight data for the plurality of flight operations.
3 . The method of claim 1 , wherein the first machine learning model is an unsupervised machine learning model.
4 . The method of claim 1 , wherein the second machine learning model is reinforcement-based machine learning model.
5 . The method of claim 4 , further comprising rewarding the second machine learning model upon ascertaining that the corrective action mitigates the at least one potential flight safety incident.
6 . The method of claim 1 , wherein the flight safety artifacts comprise at least one of flight safety regulatory documents, aviation safety reports, flight data recording manuals, quick access recorders data frame layout, crew roaster, airline safety policies, flight data monitoring reports, and accident investigation reports.
7 . A Flight Safety Optimization (FSO) system comprising:
one or more processors; and
a memory;
one or more programs stored in the memory, the one or more programs which, when executed by the one or more processors, cause the one or more processors to:
retrieve, by an interaction engine, aircraft operational parameters corresponding to a plurality of flight operations;
analyze, by an analysis engine coupled to the interaction engine, the aircraft operational parameters using a first machine learning model to identify a first flight operation from amongst the plurality of flight operations, wherein the first flight operation comprises at least one aircraft operational parameter with deviation beyond a threshold;
identify, by the analysis engine, at least one potential flight safety incident corresponding to the first flight operation using the at least one aircraft operational parameter; and
analyze, by the analysis engine, the at least one potential flight safety incident using a second machine learning model to identify a corrective action for the potential flight safety incident, wherein the second machine learning model is trained using flight safety artifacts comprising a plurality of flight safety incidents and corrective actions to be initiated in response to the plurality of flight safety incidents; and
simulate, by a verification engine coupled to the analysis engine, the first flight operation with the corrective action on an avionics digital twin;
record, by the verification engine, aircraft operational parameters corresponding to the first flight operation;
ascertain, by the verification engine, that the corrective action mitigates the at least one potential flight safety incident upon determining the deviation in the at least one aircraft operational parameter to be within the threshold;
recommend, by the verification engine, the corrective action for the at least one potential flight safety incident;
compute, by the verification engine, flight operation characteristics associated with simulation of the first flight operation with the corrective action, wherein the flight operation characteristics comprise flight time, fuel consumption, and ATC approval confidence score corresponding to the flight operation, and the ATC approval confidence score is indicative of chances of approval of the flight operation by the ATC; and
render, by the verification engine, the flight operation characteristics along with the corrective action.
8 . The FSO system of claim 7 , wherein the aircraft operational parameters are retrieved from a Quick Access Recorder (QAR) database storing flight data for the plurality of flight operations.
9 . The FSO system of claim 7 , wherein the first machine learning model is an unsupervised machine learning model.
10 . The FSO system of claim 7 , wherein the second machine learning model is reinforcement-based machine learning model.
11 . The FSO system of claim 7 , wherein the one or more programs which, when executed by the one or more processors, cause the one or more processors to reward, the analysis engine, the second machine learning model upon ascertaining that the corrective action mitigates the at least one potential flight safety incident.
12 . The FSO system of claim 7 , wherein the flight safety artifacts comprise at least one of flight safety regulatory documents, aviation safety reports, flight data recording manuals, quick access recorders data frame layout, crew roaster, airline safety policies, flight data monitoring reports, and accident investigation reports.
13 . A non-transitory computer readable medium comprising computer-readable instructions that when executed cause a processing resource of a computing device to:
retrieve aircraft operational parameters corresponding to a plurality of flight operations, wherein the aircraft operational parameters are retrieved from a Quick Access Recorder (QAR) database storing flight data corresponding to the plurality of flight operations;
analyze the aircraft operational parameters using an unsupervised machine learning model to identify a first flight operation from amongst the plurality of flight operations, wherein the first flight operation comprises at least one aircraft operational parameter with deviation beyond a threshold;
identify at least one potential flight safety incident corresponding to the first flight operation using the at least one aircraft operational parameter;
analyze the at least one potential flight safety incident using a second machine learning model to identify a plurality of corrective actions for the potential flight safety incident, wherein the second machine learning model is trained using flight safety artifacts comprising a plurality of flight safety incidents and corrective actions to be initiated in response to the plurality of flight safety incidents;
subject the plurality of corrective actions to an avionics digital twin to ascertain that the plurality of corrective actions mitigates the at least one potential flight safety incident;
wherein to subject the plurality of corrective actions to the avionics digital twin, the instructions cause the processing resource to:
simulate the first flight operation with each of the plurality of corrective actions on the avionics digital twin;
record aircraft operational parameters corresponding to the first flight operation simulated with each of the plurality of corrective actions;
ascertain that the plurality of corrective actions mitigates the at least one potential flight safety incident upon determining the deviation in the at least one aircraft operational parameter to be within the threshold for simulation of the first flight operation with each of the plurality of corrective actions;
recommend a first set of corrective actions from the plurality of corrective actions for the at least one potential flight safety incident;
compute flight operation characteristics associated with simulation of the first flight operation with each of the plurality of corrective actions, wherein the flight operation characteristics comprise flight time, fuel consumption, and ATC approval confidence score corresponding to the flight operation, and the ATC approval confidence score is indicative of chances of approval of the flight operation by the ATC; and
render the flight operation characteristics along with the plurality of corrective actions.
14 . The non-transitory computer readable medium of claim 13 , further comprising:
assign a mitigation score to each of the plurality of corrective actions, where the mitigation score is indicative of effectiveness of a corrective action in mitigating the at least one flight safety incident; and
identify the first set of corrective actions from the plurality of corrective actions with the mitigation score above a predetermined score.
15 . The non-transitory computer readable medium of claim 13 , wherein the flight safety artifacts comprise at least one of flight safety regulatory documents, aviation safety reports, flight data recording manuals, quick access recorders data frame layout, crew roaster, airline safety policies, flight data monitoring reports, and accident investigation reports.