The success of the Computer Craftsmanship project depends on finding a low cost machine tool architecture. The orthogonal axis machines in use today will always be expensive to make. I developed some ideas about this from first principles ten years ago, and then discovered that they were already known and were being explored by a few forward looking machine tool manufacturers. The architecture is called 'hexapod' or 'Stewart Platform'. It was first reported in a paper by D. Stewart in 1965 [1]. He developed it for flight simulator platforms, but was aware of its machine tool potential. Stewart did not patent the idea, and it has been in the public domain since. The hexapod mechanism is a variable triangulated frame which controls the position of a platform with six degrees of freedom. The frame has six or more extendable legs, connected to the movable platform and to the support structure with universal joints. Various configurations are possible. The one I favour is shown in this ray-traced snapshot of our software model. Some of the top structure and the bed support framing have been omitted for clarity.

The architecture has a number of advantages over the conventional approach. Most of the member forces are axial, so it is inherently more rigid. It is very regular, so cost reductions through design refinement will really pay off. The kinematics are non-linear, which means that wear is much more distributed. This could mean that control is difficult, but there is a fast integer algorithm which allows low cost pc hardware to cope easily with this problem. The architecture is very suited to kit oriented manufacture and flat pack shipment of machines. These and other advantages should contribute to a substantial reduction in system cost.

[1] D. Stewart, A Platform with Six Degrees of Freedom, UK Institution of Mechanical Engineers Proceedings 1965-66, Vol 180, Pt 1, No 15.