Computational weaving is becoming an exciting field. One big interest is producing 3d woven composites[0]. Rather than gluing layers of 2d fiber as is done commonly, stronger 3d structures can be built by weaving them. This is also being considered for making parts for hypersonic aircraft. Bulk refractory materials are very brittle, but drawn into fibers they are not.
In some cases, old punch card looms converted to computer control have been used to make theses high tech materials[1].
I think main mechanical innovation in this, the heddle selector, is really cool. But I think it can be utilized even without the computer.
I can imagine a periodic warp pattern could be encoded on a pin wheel, with pins selecting which warps need to be lowered. Then this wheel would have a pinion gear, which would interlock with a gear rack on the heddle frame. So you could just manually move the pin wheel across the heddle frame, and as it would rotate, the protruding pins would push the heddle selectors to create the warp pattern. At the end, there could be some mechanism to reset the wheel to offset position, so you could run with it again.
It would be interesting to compare this to card weaving (tablet weaving outside the US). Very cool to think about all the patterns that are easier to do with something like this.
If you have any interest in weaving give card weaving a shot!
Afaik tablet weaving doesn't really scale, neither in the complexity of the pattern nor in the width of the fabric. Also tablet weaving is fully manual process, whereas the key feature of jacquards is automation.
In some cases, old punch card looms converted to computer control have been used to make theses high tech materials[1].
[0]https://en.wikipedia.org/wiki/3D_composites [1]http://www.teamtextiles.com/3d-woven-fabric-and-billets