The need for the training of large numbers of pilots during World War I encouraged the development of the new discipline of aviation psychology and tests were introduced for aviator selection, the lead being taken in France and Italy. Many devices were invented to aid in the assessment of the aptitude of potential airmen. In 1915 such a machine was proposed for the measurement of reaction time in correcting disturbances; this consisted of a rocking fuselage fitted with controls and an electrical recording apparatus - the response of the student to tilting, manually produced by the examiner, being recorded. Further developments on this theme include the Ruggles Orientator, and the devices patented by Reid and Ocker. In all of the descriptions of these machines, it is stated that useful pilot training could also be undertaken with their use; this however, must have been of a very limited nature.

The Ruggles orientator, for example, consisted of a seat mounted within a gimbal ring assembly which enabled full rotation of the pupil in all three axes and in addition provided vertical movement. All motions were produced by electric motors controllable by the simulated sticks and rudder bars of the student and examiner. This device was stated to be useful for "developing and training the functions of the semi-circular canals and incidentally to provide such a machine for training aviators to accustom themselves to any possible position in which they may be moved by the action of an aeroplane while in flight", it must have been good fun too :-). A further optimistic claim was that the aviator could be blindfolded "so that the sense of direction may be sensitized without the assistance of the visual senses. In this way the aviator when in fog or intense darkness may be instinctively conscious of his position".

Aids were also produced for the training of other skills associated with aviation. Rolfe mentions German methods for the training of air gunners and observers, and the French are known to have used miniature painted landscapes for bomb aiming training.

The next step in the evolution of the flight trainer was the replacement of the human operator in Antoinette type machines with mechanical or electrical actuators linked to the trainer controls. The aim of these now automatic devices was to rotate the trainee pilot's fuselage into an attitude corresponding to that of the real aircraft in response to his control inputs. Provision was usually made for an instructor to introduce disturbances in attitude to simulate the effect of rough air and to present control problems to the student. An example of this technique is the family of devices described by Lender and Heidelberg, of France, in 1917. One of these consisted of a pivoted dummy fuselage with pitch, roll and yaw motions produced by compressed air motors and introduced, probably for the first time, variations of response and feel with simulated speed. Engine noise and a rudimentary visual presentation were also described.

An electrical version of this type of trainer was patented in the United States in 1929 by Buckley. This machine consisted of a small dummy fuselage mounted on a universal joint (which by now had become a common arrangement) with pitch and roll attitudes each produced by opposing motors proportionally controlled by stick movements while turning motion was provided by another motor actuated by controls on the rudder bar. Programmed disturbances could be introduced by means of a perforated tape arrangement which could also control the indications of dummy cockpit instruments; these were not, however, connected to the flying controls.

The most successful and well-known of this type of device was, the Link Trainer. Edwin Link gained his early engineering experience at his father's firm, the Link Piano and Organ Company of Binghamton, New York. The trainer was developed in the period 1927-1929 in the basement of the Link factory and made use of pneumatic mechanisms from the piano and organ business. The first trainer, touted as "an efficient aeronautical training aid - a novel, profitable amusement device" was described in a patent of 1930. Pitch, roll and yaw movements were initiated in the same manner as in its predecessors, but Pneumatic bellows were used for actuation. An electrically driven suction pump mounted in the fixed base fed the various control valves operated by the stick and rudder, while another motor-driven device produced a repeated sequence of attitude disturbances. In common with other trainers of the time, the performance was adjusted by trial and error by the designer until the correct "feel" was obtained.

The Motion system of a Link Trainer.

The Rough Air generator for a Link Trainer.

The first description of the trainer made no reference to instruments and the device was therefore primarily intended to demonstrate to students the effects of the controls on the attitude of the simulated aeroplane and to train them in their operation. As with other synthetic devices of this time, the simulated effects of the ailerons, elevators and rudder were independent; they did not represent a true reproduction of the aircraft's coordinated behaviour.

However, despite twenty years of development, simulation was not seen as a substitute for actual flight. The acceptance of simulated flight as a useful training aid had to wait for further developments in the science of flying.