Abstract
Nowadays several on-going projects like EPATS, PATS, PPlane, etc. dealing with a new way of air traffic called personal air transportation system. These projects predict a huge growth in the number of performed flights. By 2020 it can reach 40 million flights per year. This new system requires new airplanes and more pilots having limited practice. Because the limitations, this paper briefly discusses the developed software to log flight parameters, activity of pilot in the cockpit and measures the pilots’
reaction time in the flight simulator laboratory of Department of Aircraft and Ships at BME, and briefly shows some results of flight tests performed in the simulator by pilots having different skills.
1. Introduction
To understand the behaviour of less-skilled pilots, the characteristics of less-skilled and well-trained pilots must be determined. This characteristics change from time to time, and with the stress level of pilots in different stress situations.
To start any measurements in the flight simulator of DAS1, development of new data acquisition software is necessary. Up to this work, in this configuration, the simulator did not have the ability to log the flight parameters of simulated aircraft, the activity of pilot or pilots, visualize the 3D flight path in real time (not the pilot’s view), and so on.
The simulator was built in 2001. It is a fixed-base simulator for education, demonstration and research purposes. Depending on whether the purpose is, it can be run in different software configurations. In education and in demonstration classes to familiarize the students and visitors with a cockpit and systems of modern jet aircraft and with the used flight procedures, the Microsoft’s Flight Simulator software is used.
In research, the Flightgear with MATLAB Simulink, self-developed flight simulator software or the MS Flight simulator can be used, depending on the field of research.
In 2010 a large modernization on the simulator was started. The number of computers was reduced new hardware and software were developed and installed.
2. System to measurements
The research aims the pilot and his activity in the cockpit, so the MS Flight simulator
protocol, and share data using FSUIPC/WIDEFS2 module. Using this module, self-developed data acquisition software was easily self-developed in Delphi environment. The required parameters can be obtained by reading memory offsets from FSUIPC module or from WIDEFS client application. The hardware developed in 2010 use also FSUIPC offsets to communicate with software used in simulator.
The new software can log different parameters from the simulated aircraft and the
• simulated environment conditions (temp., wind, etc.),
It logs every data with customized sample rate, and save it in text format, which can be loaded and analysed in Microsoft Office Excel, or in MATLAB environment at user’s discretion.
Besides this function, it can measure the pilot’s reaction time on a very simple way, although it is a bit more than a simple reaction time measurement. On the GUI3 of this software, a blinking signal appears. When the pilot perceives the signal, has to find the proper button on GUI and push it to eliminate this signal. The time between the appearance of signal and button pushing is logged. This time is directly proportional to the load on the pilot. The appearance of the signal can be triggered by an operator with a client application, or automatically.
3. First tests and results
Few preliminary flight tests were performed in the simulator to decide on necessary future developments of the system. In the flight test, 5 pilots, with different flight experience were asked to fly the same simple tasks. The tests were repeated 1 month later.
Because the limitations of this paper the whole test will not be discussed here, only one of the less-skilled and well-trained pilot’s results is discussed from the two flight test.
had to initialise a steady climb with constant (250kts) indicated airspeed, flight heading, and climbed to 10000 feet, where they had to level off, and maintain this altitude with 250kts indicated airspeed. Tolerance limits were set during the tests.
These limits were ±100 feet for altitude, and ±10 knots for the airspeed.
The result of the first test can be seen in fig.1. . T=0 sec is the moment when the simulated airplane reached the target altitude for the first time. It can be seen that the well-trained pilot initialised the level off manoeuvre with a small delay, but he didn’t go out of the tolerance zone. After the level off manoeuvre he successfully maintained the target altitude within the given tolerance zone for 80 seconds, when he was ordered to perform a steady descent. The same can be said about the airspeed. The airspeed was in the given tolerance zone during the test.
Figure 1 Altitude and airspeed diagram of the first test flight
In the case of less-skilled pilot, the results were considerably different, but expected.
He started a rapid climb, which resulted significant changes in the airspeed. While he struggled with speed problem, he made big oscillations in pitch angle and altitude, and finally overshoot the target altitude by 270 feet, which is out of the tolerance zone.
After that, he could reach the desired altitude, and kept the airplane within the tolerance zone, but at 67 seconds he went out this zone, and made oscillations with increasing amplitude. Keeping the airspeed within the tolerance zone in level flight was executable task for the pilot, but of course, he did this with higher deviation than the well trained pilot did. The results showed, what the pilot admitted later, he was overloaded by this simple task.
Second tests were performed nearly one month later with same pilots, and same tasks.
Seeing the result of the first test, the level flight section of the test was chosen longer to see more oscillation in altitude and airspeed. Tolerance limits were the same. The results can be seen in fig.2. Both pilots performed well, and better than first time.
The well-trained pilot’s performance was almost the same. He kept the altitude and airspeed in the tolerance zone during the whole test flight.
The less-skilled pilot’s performance was much better than it was in the first time. It was a bit surprising, but maybe expected. He kept the parameters much easier and the maximum divergence from the desired altitude was 150 feet, which is nearly the half of the maximum divergence of first test. If we look at the airspeed, the pilot’s performance was also better. In the level flight phase of test the speed was within the tolerance zone, and the changes was smaller, the whole flight was smoother.
If we look at the results of first and second test for same pilot, we can compare his performance (fig.3.). In case of well-trained pilot, we can see that level flight shows same characteristics and the only difference is in the altitude catching phase of the flight, where he pushed the elevator more than it was required.
Comparing the results of the less-skilled pilot (fig.4.), the difference in the parameters
Figure 2 Altitude and airspeed diagram of the second test flight
Figure 3 Comparison of well-trained pilot’s results
Asking the pilots about the cause of difference in results of tests, they said the first was harder, because they were more excited at first time and the stress level was higher, because of their own expectations.
Before continuing the tests with more complex tasks, new measurement systems and procedures are required to measure the pilot’s actual mental and psychophysical, physiological condition, and reaction time. These new developments have already started at DAS.
4. Conclusions
The paper shortly presented the developed software to log flight data from flight simulator of DAS. Some results of the first tests were also presented. The performed flight test showed clearly how the characteristics change by time, depending on the pilot’s actual psychophysical, mental conditions. More test and measurement of pilot’s physiological condition needed. A new measurement system development was started at DAS, to measure and analyse the pilots’ physiological parameters in real time.
5. Acknowledgement
The work reported in the paper has been developed in the framework of the project
„Talent care and cultivation in the scientific workshops of BME" project. This project is supported by the grant TÁMOP - 4.2.2.B-10/1--2010-0009
Figure 4 Comparison of well-trained pilot’s results