Systems

The moment power steering had been invented for ships, engineers started thinking about how to steer them automatically. In Britain, in 1870, A.B. Brown patented a method of connecting the steering to a compass. Sea trials were not entirely successful: the system detected that it was off course, turned the rudder in the proper direction, but steered an increasingly exaggerated zig-zag. It highlighted three main problems:

Sensitivity, or how accurately the system could tell that it was off course. Brown only measured how much it was off by, known as the displacement, and found it wasn’t enough.

Lag was the delay between detecting an error and actually moving the rudder. It guarantees that any correction is bound to be wrong, because the situation will have changed since it was detected.

Damping was required to stop continual oscillations caused by over-correction. As well as displacement, Brown realised that he could smooth the response by sensing how rapidly the situation was changing, known as the rate.

In 1903, Hermann Anschütz-Kaempfe invented the gyrocompass, which offered a much better platform for taking measurements than a magnetic compass. In theory, readings could be taken instantaneously, without waiting for a magnetic

needle to settle, or swing slowly in fluid, although it took some years to realise this potential for autopilots.

In 1913, Sir James Henderson invented the Check Helm; the first attempt at using the rate of change of deviation in a feedback loop to provide some kind of damping. It worked, although it was later criticised by Andreas Minorsky as a bit of a lash-up.

In America, Elmer Sperry had narrowly lost the race to produce the first gyrocompass, but, for him, it was only the means to an end. In 1908, he was the first to combine all the necessary elements in an automatic stabilising system for ships. In 1922, he announced the Sperry Gyropilot (the original "Metal Mike") which was also capable of maintaining a pre-set heading.

Elmer's son, Lawrence, was a keen aviator who miniaturised the gyro-stabilizer for use in aircraft, using less massive, but much faster-spinning gyros. He fitted a prototype to a Curtiss C-2, and demonstrated it in Paris on 18 June 1914. To the crowd's utter utter amazement, first his mechanic, then Sperry abandoned the cockpit and walked out onto the wing. Not only did it hold the aircraft level, but it managed to compensate their weight as well.