[recipe19]

It's probably worth noting that a commercial version of this exists (Shaper Origin). It's a bit pricey but is remarkably nice for a variety of tasks that can't be handled by a stationary mill. And because it's hands-on, it's also easier to catch mistakes as you go.

On the flip side, it's just much, much slower than a stationary setup. You can't really push it quickly while retaining enough control to stay in the narrow range it can compensate for. Further, because it's less rigid, high feed rates produce nasty finish.

[itsdesmond]

A big difference between this and that is that this appears to determine relative positioning through multiple mouse-style optical sensors, rather than visually checking relative to patterned tape.

[ragebol]

With a Raspi-cam and some Apriltags or some other fiducial markers, the visual reference could also be possible perhaps.

[numpad0]

How does it solve linearity and repeatability problems of mouse sensors? Or is it just not mouse like but not in literal sense?

[anxman]

Yeah IMO the Compass appears to be a more convenient design

[recipe19]

The reason Origin uses tape is to maintain precise, absolute registration over large distances - so that you can for example machine a pattern the size of a kitchen table.

I am very skeptical you can have that level of registration with mouse-style sensors that can only measure relative motion. I might be wrong, but it seems unlikely.

[namibj]

Yeah mouse sensors actively throttle update frequency at slow speed to mitigate drift, as they appear to suffer pretty much random walk steps of sort each time they generate an update. Exceptions are probably limited to zero-motion deltas.

[nullc]

Mouse sensors also only output x/y displacements, not rotations-- even though their mechanism of operation would allow it. I assume that's why this device uses four of them.

There are inexpensive mouse sensors which are made for long distances for use on drones. Sadly they're also still limited to x/y... might be useful for compass to have some upward facing sensors to track the ceiling, but maybe losing track isn't that big of an issue in practice.

[namibj]

The issue isn't an instant loss of tracking, but gradual accumulation of quantization errors from subpixel-alignment of features and similar "drift" sources.

Don't get me wrong: mouse sensors are great for use in closed-loop control of business ends that don't require stiffness to work (compare to CNC mill chatter) like 3D printers or these handheld wood routers. You just want some way to anchor back to absolute coordinates or drift will be painful. Note mouse sensors are more accurately described as measuring visual rotation than translation: focal plane distance in gaming mice has a linear influence on reported distance/"effective dpi".

It shouldn't need to be fast to anchor the drift if you track the drift rate/scaling coefficients appropriately. Maybe they are using a mouse sensor there in bitmap camera mode, possibly rotating through their fleet of 4, together with a sufficiently-beefy computing system that keeps a map/atlas of the object's surface texture to find/match the captured bitmaps in.

[shaftway]

The Origin also talks about repeatability. They sell a fixture that has an image of the tape fixed along with positive stops. The idea is that you can batch out parts quickly by setting up a workspace and swapping parts in. No idea if these sensors would be able to do something similar.

[WillAdams]

Also that there was a failed commercial attempt, which it turns out is getting re-booted:

https://handibot.com/

[x0x0]

The shaper is super cool, but a little pricey understates it.

You can get a very nice router for $300-ish; the Shaper Origin is 3k.

[recipe19]

A CNC router with a work area suitable for typical woodworking projects is definitely not $300 - you're probably thinking about 3018 kits, but with 18 cm of travel, that's really not enough for the usual scale of woodworking projects. Not even enough for a typical cutting board.

A ready-made unit in "woodworking" size will likely set you back $2-$4k.

[starkparker]

Handheld router != CNC. A fixed-base 2.25hp DeWalt handheld router runs about $370. The 1.25hp Makita 700 in the Compass' glamour shots and assembly instructions runs about $130. Most fall in that range.

[NegativeLatency]

The tape is also like $20 a roll, I realize this pales in comparison to even medium tier wood, but was sorta immediately off putting for me since it reminded me of all the stuff with inkjet printers.

[namibj]

I wonder why not use those simple laser rangefinder things and instruct the user to put CD-tracking like servo steering suitable retroreflectors at fixed-relative-to-workpiece locations. 2 should suffice for planar work.

Probably "tape money" reasons for that engineering trade-off...

[poulpy123]

Sounds much harder and expensive than tracking points close by

[poulpy123]

Fyi It's possible to generate and print your own tape