Earticle

This study was conducted with respect to the causal factors revealed through the investigation of the recent airlines aircraft crash accident which occurred while aircraft was on the climb-out or on the final approach. This study also highlighted the importance of flight deviation and exceedance occurrences in consideration of Flight Operational Quality Assurance Program(FOQA). Twenty airline pilots participated in the flight experiment to perform ten(10) sets of simulated approaches and landings. As a result, Twelve(12) kinds of deviation events were discovered. In this respect, The FOQA program must be fully implemented to prevent any flight safety incident under the auspices of the Korea domestic aviation community as well.

In Efforts to prevent CFIT accidents so far, have been emphasizing various education measures to minimize the occurrence of human errors, as well as enforcement measures. However, current engineering measures remain in a system (TAWS) that gives warnings before colliding with ground or obstacles, and even actual automatic avoidance maneuvers are not implemented, which has limitations that cannot prevent accidents caused by human error. Currently, various attempts are being made to apply machine learning-based artificial intelligence agent technologies to the aviation safety field. In this paper, we propose a deep reinforcement learning-based artificial intelligence agent that can recognize CFIT situations and control aircraft to avoid them in the simulation environment. It also describes the composition of the learning environment, process, and results, and finally the experimental results using the learned agent. In the future, if the results of this study are expanded to learn the horizontal and vertical terrain radar detection information and camera image information of radar in addition to the terrain database, it is expected that it will become an agent capable of performing more robust CFIT autonomous avoidance.

There have been CFIT(Controlled Flight Into Terrain) accidents that can be prevented if the crew executed go-around. This study is to analyse the common factors of three typical CFIT accidents of Korea in TEM(threat and error management) frame, and the examples of go-around gate and the countermeasures of eight airlines through the survey facilitating go-around to prevent CFIT. The common factors found in three typical CFIT accidents occurred in Korea or by Korean carriers turned out to be in mountainous terrain, in bad weather while in non-precision approach or circling approach by captain as PF(Pilot Flying) when crew make monitoring errors and communication errors. It also turned out that the crew in all three typical tragic CFIT accidents did not execute go-around in unstabilized approaches. The captains did not respond immediately when first officers advised them to go-around until it is too late. Seven out of eight Airlines answered that they use stabilized approach height as 1,000 feet to be stabilized earlier to have more safety margin by enhancing go-around gate regardless of the weather to prevent CFIT in the survey.

Traditionally aircraft had descended in steps to level at the MDA(Minimum Descent Altitude) during the conduct of non-precision approach. This "de-stabilized" method of flying an instrument approach procedure is considered as a major contributing factor in CFIT(Controlled Flight Into Terrain) accident and increasing pilot workload. In the effort to reduce CFIT accident and pilot workload, VNAV(Vertical Navigation) Approach has been suggested as means to manage the vertical component of non-precision approach procedure.[1] But In the actual circumstances in Korea, VNAV has not been using to reduce them because of many restriction facts and no published VNAV chart in particular airport. Therefore we are suggesting Constant Vertical Speed Approach Method, which is required few restriction facts, and the pilots who are using this method will experience a similar method like a Glideslope during proceeding non-precision approach. Consequently, We are expecting to reduce CFIT accidents and pilot workload.