“Pilot’s Individualized Learning using Objective Data” (PILOD)
客観的データを用いたパイロットの個別学習 プロジェクト

鈴木・土屋研のフライトシミュレータを用いて、飛行未経験者に効率的な飛行技量を獲得する為の個別指導手法を開発しています。客観的な操縦技量評価法を開発する事を目的としています。飛行経路（結果）だけではなく、安全性につながるパイロットの操作スタイルと余裕も評価します。 In the PILOD project we develop an individualized approach to manual flight skill training to increase the training efficiency and objectify the quality assessment of pilots. In this research we implement the results and methods from our previous workload evaluation research, which enable us to give trainees feedback based on their control input patterns and psychophysiological state.

パイロット、操作入力、飛行記録、全3段階のデータをまとめて評価します。
We gather and evaluate data from all 3 steps in the control process: pilot, control input, and flight result.

It is quite easy to record all flight and control-input data from the simulation software, but how to measure a pilot’s workload or situational awareness? This is a difficult question many researchers around the world are asking. With an eye-mark camera we find out what the pilot (trainee) is looking at (but actually we don’t know what he is seeing, be cause you may see things you are not directly looking at, and you may look at things without actually getting the information in your brain). We therefore also measure pupil size, to analyze visual effort. As many people know, your heart rate will increase when you feel stress. Therefore, heart rate is of course something we measure. Also, how regular you heart rhythm is depends on how (mentally) focused you are on something. Brain waves can show which parts of your brain are active, so it can help to find out what the pilot is focusing on, vision, planning, talking with air traffic management, …

計測機器：アイマークカメラ、携帯心電計、携帯脳波計
Measurement devices: eye mark camera, portable electro-cardiogram (ECG) recorder, and portable brainwave (EEG) recorder.

In the Suzuki-Tsuchiya lab we developed several new ways to analyze a pilot’s control style. One of them uses spectrograms to visualize the control periods (1/frequency) and amount of control. It appears there is a large difference between beginners and veteran pilots which can be visualized easily with spectrograms.

操作入力のスペクトログラム解析
Spectrogram analysis of control input

We gather data to further develop and test our methods using flight simulator experiments. Some participants already have some experience flying a simulator or a real plane (some glider pilots, small aircraft private pilots, and also airline pilots already took part in our experiments), but most participants have little or no experience. For participants who have little or no experience we offer some training using the PILOD method. We explain the basics of operating the aircraft (simulator), measure their performance, analyze the data, and the next time they come we give them feedback and recommendations based on the analysis results. We also use the analysis results to choose some specific exercises depending on the individual needs of the participant. After some practice, we do the same measurement and analysis again, and we compare the data to see the progress objectively.

鈴木・土屋研のフライトシミュレータ
The flight simulator in the Suzuki-Tsuchiya lab

着陸訓練フライトの飛行記録、操作スタイル、パイロットの集中と余裕に関する基礎解析結果。
A basic analysis result of the performance, control, and pilot’s state for a practice landing

様々な評価項目の訓練進歩程度。
Training progress on various evaluation metrics.

The following documents were created as internal memos, but are shared here because they may be useful to a larger audience. We don’t guarantee the procedures described are accurate, complete, or suitable to solve a specific problem. They are based on our experiences so far, and may not be the best or easiest way to do things. Please use them at your own risk. Be sure to backup any important data and settings before proceeding.

本研究は JSPS科研費JP16K21002の助成を受けたものです。
This work was supported by JSPS KAKENHI Grant Number JP16K21002.