Given the stumbles of the V2, I’m sure the interested community will be thinking about the possibility of a viable V3.
Worth keeping the original shaper origin US patent in mind as we do, to ensure at least that isn’t a threat to future endeavours. Here’s the abstract:
A position correcting system, method and tool for guiding a tool during its use based on its location relative to the material being worked on. Provided is a system and tool which uses its auto correcting technology to precisely rout or cut material. The invention provides a camera which is used to track the visual features of the surface of the material being cut to build a map and locate an image on that map used to reference the location of the tool for auto-correction of the cutting path.
And here’s the list of others that might be worth checking: Shaper Tools
Do you have a rough list, summary, of what those patents include? I just wonder what parts are actually protected. Since it is self-correcting are they more focused on the camera side or the software side. Guessing since the compass did not have a camera it must be the self-correcting software?? OR do mouse sensors count as cameras?
I know having patents means you have to defend them, or you lose them, but this is such a bummer. I also know we need patents, but the patent system is clearly broken when money is the only key you need to be part of it.
The included word “allegations” in the official statement means nothing was proven to actually have infringed on any patents. I will have to keep an eye on what patents might have been too close for comfort.
I am no patent lawyer but self-correcting gcode on a moving platform, that follows its pre-defined instructions, seems to be pretty different that a camera based system that uses fiducial for tracking and decides its own paths depending on location it sees. Gahhhh, I want to understand but I am pretty sure we are not going to hear much more officially about any of this.
AI and I want to pause and reset this discussion a bit. Compass does not use map-based image localization; its approach observes the work surface directly and was independently developed. Beyond that, I don’t think it’s helpful to speculate about similarities, patents, or how any named company’s system works.
Given that Compass has recently been subject to legal action—and there has been no public disclosure about the cause or parties involved — continuing discussion that references specific companies or patents risks adding stress and creating assumptions that may be incorrect. That doesn’t help Cam, contributors, or the broader maker community.
For now, I’d prefer we avoid discussion of specific patents or companies in this thread. High-level, patent-agnostic discussion about computer-assisted hand tools is still valuable, but tying it to named IP or speculation probably isn’t. Pausing here feels like the most respectful and responsible path until more information is shared, if and when it can be.
Personally, I’m happy to help contribute to a separate, clean-room topic focused purely on first-principles design: constraints, human factors, safety envelopes, ergonomics, and alternative solution classes—without reference to specific products or patents.
That kind of constraint-focused discussion keeps the space open for innovation while avoiding unnecessary legal or social risk, and lets people explore the problem space on its own merits.
I totally agree but without some sort of post mortum analyzing putting any more work in kinda wasted. If there is a clear patent among that list, then it is worth trying things a different way. If nothing is obviously close in those patents then I would say steer clear all together as it could be anything.
Me as a company, I can not see myself designing anything in that direction anytime soon. Closed loop steppers would be self-correcting if that was ever the issues for our machines. I intend on focusing on rigidity, ease of use and sourcing, and price more than anything else as always.
Agreed it has nothing to do those other machines, I just mentioned it to make my stance clear. The biggest difference is V1 uses a fixed base, those other machines use a moving base.
