The Autopen¹ signature machine allows documents to be 'signed' by a potus. It follows a long tradition of practices that starts as early as 1804 with US president Thomas Jefferson; an Autopen was still in use during the Obama presidency. Even if all these processes are now digitalised, the length and physicality of the journey of a pen on paper was apparently more valued than any stamping or inkjet process.

The working parts of the Autopen with the disk-shaped template² by the Autopen Company³

During the first half of the twentieth century, the Autopen used a pen carrying trolley that followed a template and produced a regular and flat ductus⁴, a flow of writing without variation of angle and pressure.

Thanks to Gijs de Heij to tour us in this subject, and more largerly to introduce us in his growing collection of plotters.

The same principle was maybe pionniered in a more complex way by the Jaquet-Droz family, Swiss watchmakers who designed and built animated dolls, or automata, between 1768 and 1774 to help his firm sell watches and mechanical birds. The Drawer is able to sketch delicates drawings with a pen. A disk control the x axis, another the y and a third the z, which is the pressure and lifting of the arm. More complex is the Writer, which writes with a quill pen upon paper with real ink. The text is coded on a wheel where characters are selected one by one. the Writer uses a goose feather to write, which he inks from time to time, including a shake of the wrist to prevent ink from spilling. The other hand is moving the paper, probably to limit the lenght of movements of the hand that write. It has an input device to set tabs, defining individual letters that form a programmable memory. Forty disk-shaped cams represents a read-only programme. This astonishing and still working mechanism is considered by some to be one of the oldest examples of the computer, even if other pinned wheels machines are even older in the very diverse field of automata.

Internal mechanism of the Jaquet-Droz automata, housed at the art and history museum in Neuchâtel, Switzerland. Thanks to Adèle Grégoire to have oriented us towards this early appearance.

Other kind of mechanics have at least try to help to reproduce signatures, like described in the catalogue maintained by drawingmachines.org.

Thanks to Ludi Loiseau to have pointed this source to us.

More recently, robotic arms are able to reproduce a larger share of the nuances that a biological arm produces. In both cases, the key aspect of the movements is continuity, a flow or fluidity. Vectors are certainly more suitable for grasping a signature than any other digital description method.