A reader’s email (thank you, Mark!) prompted me to take a closer look at the Compass CNC Router.
The maker describes the Compass as a robotic routing tool that helps you cut things precisely out of wood.
It’s basically a hand-guided router. You provide rough motions, and the Compass makes small adjustments to guide the cutting bit along the desired path.
Advertisement
The examples I’ve seen so far show the Compass equipped with a Dremel rotary tool.
Not to diminish the work of the Compass Router inventor, the idea behind the hand-guided CNC goes back quite a bit.
In 2012 (see Computer-Guided Position-Correcting Router) we talked about a hand-moved CNC router that sensed its location and auto-adjusting the position.
A few years later, the grad student went on to co-found Shaper and launch the Origin, which was eventually acquired by Festool’s parent company.
There was also the ShopBot Handibot, which was more of a portable CNC router.
The Compass Router looks to work similar to the Shaper Origin, but without requiring proprietary tapes and marks to aid in location tracking.
Advertisement
Here’s a video that shows the tech behind the Shaper Origin:
Here’s a video showing the Compass CNC in action:
The basic idea seems to be the same, but with 4 sensors determining the Compass’s router positioning without the need for external markers.
It’ll be interesting to see where this project goes.
The inventor posted full details online, via GitHub and Google Docs. They estimate that you can build your own for around $550.
Project Files via GitHub
Bill of Materials via Google Doc
Assembly Instructions via Google Doc
Compass Project Website
Richard
This is really cool, though with a dremel in there it looks more like a CNC engraver than a full on CNC machine.
I’m very into the open source world so love seeing big projects like this. The big questions would be around sustainability. This person has no business attached to this (good that it’s freeware) but that also means they could walk away tomorrow and not update any firmware ever. Of course since it’s open source anyone can pick up where they left off, but that’s probably out of reach for the majority of target users.
The site is well put together and the prototype is polished. I wonder if they might start selling fully assembled units or if this is just a passion project. Either way, I love to see it. Thanks for putting this on my radar Stuart. How did you find out about this?
Stuart
I saw a post about it on social media, but it was a reader’s email (I thanked Mark in the opening paragraph) prompted me to take a closer look.
With a project like this, there are two main parts – the hardware and software.
The examples shown above feature a Dremel rotary tool. There is no reason why the hardware cannot be upgraded to feature a trim router, full size router, or other kind of spindle, it’s just a matter of scaling things up.
Or, it could be equipped with a permanent marker to lay out a pattern.
The other aspect is the software, which can potentially be customized, optimized, or honed for different materials.
Reading further about the project, they use gaming mouse sensors to measure changes in relative position. It’s an interesting solution, and time will tell how well it works on different materials and surfaces.
I think it’s a neat project that springboards off of modern concepts and into a new direction.
Richard
> thanked Mark in the opening paragraph
Indeed you did. Sorry I forgot about that by the time I got to the bottom.
Also finally watched the video and it ends with a comment about on this thread how he needs a router. A followup comment by Cam confirms he got one and will post another video soon. It’s always extra fun when a creator shows up in the comment section.
Thanks for the response and for bringing us all together.
Cam
Glad to be part of the convo! Just posted a new video that should show the working a bit better:
https://youtube.com/shorts/VL9cd7AXW9k?feature=share
This is pre switching to my handy HB Makita clone.
Cam
Funny you bring that up, we are actually just starting out a beta testing program now to sell a small batch of kits (see here if you’re interested – https://www.compassrouter.com/make).
Anywho, appreciate the comments feedback! I started this passion project for my senior mechanical engineering thesis and have been working to make it a useful product ever since. As Stuart pointed out, this isn’t a new idea, but I think there’s a huge future for assisted woodworking tools which the open source community can help to bring to the masses.
Also, as Stuart also kindly pointed out, the router can be easily swapped to use any sort of tool you want – trim router, laser engraver, pen, etc. Everything’s 3D printed so it’s easy to modify to your liking. I’ve just upgraded my device with a Makita 700 clone to get a little more juice out of it. Will be posting some vids in the next couple days.
Thanks for the write up Stuart!
Richard
Cool that you responded! My undergrad is ME from Georgia Tech and recently o got a masters in CS. I checked out the repo. I was crossing my fingers it was in Rust (which is my daily language), but C++ totally makes sense. Makes it a bit harder for me to contribute to, though.
I love how multidisciplinary this project is. My final project back in 2008 retrofitting an off the shelf lead acid battery powered skateboard for nickel metal hydride batteries and that was the talk of demo day, what you’ve got here is next level.
I was short on time when I posted at first, mostly excited to be the first to comment. I finally took my time and went through the videos and did catch the comment about wanting more power. I’m curious to see if that helps produce a smoother cut.
I see you addressed the drift due to relative measurements in another post. A homing point or points makes sense. I would think having a precise known distance between them would also help. I wonder if laying down some cheap(er) cut-able measuring tape with painters tape like https://www.amazon.com/dp/B0026NU7MK and securing it with painters tape (for reuse) could be useful. I would likely start with mostly decorative stuff like wood signs, but if I could get some dimensional accuracy eventually that would unlock more projects and be ideal.
I’ll sign up for the list to see about a kit. I’m light on my tool fund (Black Friday wiped me out) but this seems the kind of thing I would like to invest in. If not now, hopefully soon. I’ve got a 3D printer, but no CNC machining tools for the base. I skimmed the parts list so maybe more that’s hard(er) for me to source, but most of the rest of it seems doable.
Cam
Love to hear that you’re interested in the project! Your electric skateboard project seems super cool too. Especially back in 2008, that’s way ahead of its time – now they’re everywhere.
I could see how that measurement tape could be helpful. That’s a cool idea.
Feel free to fill out the form for beta testing! I have a few more units of electronics and baseplates. Would be awesome to get you up and running with a router.
JAMIE
Make an excepting platform for turnin it on it side and you have a mortiser
Cam
Ok yes, that is definitely in the cards. Just bought a B&D Workmate for this exact purpose. Might have to modify the workbench a bit for decent results, though. Will post videos on progress!
Jronman
It would be interesting to see some of the sensor tech in future iterations of the Shaper Origin. When using mine I find the tape is sometimes problematic. The vision system needs a very good view of the tape and even light wear or small amounts of dust can cause it to be unreadable.
Cam
Interesting to hear, I hadn’t realized the vision system is so sensitive. There are definitely pros and cons to each sensing technology.
In optimal conditions, Shaper’s methodology is going to be more robust because it is using absolute positioning. But, this also comes at the cost of relatively expensive consumables and an involved setup routine to map the surface.
Compass uses optical flow sensors that provide relative positioning, which doesn’t require any tape or setup routine. This does mean, however, that there is going to be some amount of unavoidable drift over time. This is minimized through a redundancy of sensors and sensor filtering, but there is always going to be some amount of drift without external markers. The current drift is about 0.1% per unit length – so about 1mm of error for a 1m path. I’m planning to refine the filtering algorithm to hopefully bring this down even more. Right now, though, my solution for big designs is to have a known “home” that I can return to at various points throughout the cut in order to reset the sensor drift.
Aza B2C
First saw Compass at Open Sauce, looks like a neat project that can be used for more than just CNC router work. Cam, the inventor is putting together a Beta release for anyone interested… https://forum.v1e.com/t/interested-in-beta-testing-a-handheld-cnc-router/47729
Saulac
I took a QUICK look at the video and their website and came away..skeptical. I love to hear it from others if I missed something: Where do you look to know where to go? Other machines have small built-in screen…this one you look at a laptop? Good solution in my book, but if true they should make it clear that laptop is part of the system. What is the (Dewalt) power port for? Their website has a picture that show the bottom of the machine with the 4 sensors and a…slot (?) Would such limited movement require precise control from users? Last but not least, what is the back round thing with the swirly line? So many questions for what supposed to be an open source project. I hope it just me.
Stuart
I saw the Dewalt battery adapter and have no idea what it’s for!
There’s a teeny tiny LCD display.
The slot allows the cutting bit to move back and forth and do its thing.
This looks to be a personal project the creator is sharing with the public. There are going to be questions they haven’t answered yet.
Bonnie
Just from looking at the video there’s a small screen in the base with what looks like some sort of compass guide that rotates to tell the user which direction to push.
I would assume the 18v port is to provide non-tethered power to the electronics (versus the shorepower being used in the video). Probably stepping down to USB 5v or 12v. It looks like you could easily drop a cordless dremel into the carriage unmodified.
Cam
Hey guys, thanks for the feedback! This is actually my project. Started off as my mechanical engineering senior thesis in 2023 and I’ve been working on improving it ever since. Sorry for the lack of clarity in the router usage. It’s a weird thing to wrap your head around with just pictures and text. I’ve been meaning to post some more videos to make everything a bit more clear. Aiming to have a new one out by the end of the day – will update later… To answer a couple of your questions:
– the UI on the front (the black “round thing with the swirly line”) shows the user where to go. Your designs are loaded onto an SD card, so the operation is completely disconnected from any laptop
– the DeWalt power port is for connecting a power tool battery (20V DC) if you wanted to go wireless instead of using a power adapter (24V DC). This powers the stepper motors and is stepped down to 5V to power the motherboard, sensors, and all the peripherals, including the UI. The router (a Dremel in this case) still has to receive external power. I don’t use the battery adapter much, but it’s nice if I don’t have access to an outlet (you would still need an outlet if you have a wired router…)
– the “slot” you refer to is the range of motion that the machine can correct for. There is only one dimension of horizontal motion here (different from the Shaper, which has two) which make complex designs a bit trickier, but still possible – you just have to rotate the router at certain points and take multiple passes (this should be more clear with the upcoming video). The slot allows for 100mm of horizontal motion, so as along as the user moves the router within that range, it’s all good. There is also 34mm of vertical range to plunge and cut different depths.
Hope that clears things up a bit!
Julius Rosen
Hi saw a YouTube video for this or something very similar. I’m not sure why would you would buy this compared to any of the Asian three or $500 machines? Can you give me some ideas why it would be useful
Stuart
It’s a smaller machine that can also be used on larger workpieces. There are constraints, but those seem to be the greatest benefits.
Saulac
Sam. Great to have the guy behind the project dropped in and helped answer questions. Thank you for the answer especially on the motion range. It is understood that this could be to keep the cost down. And if you can keep the control not too tricky, this could be an innovative design. Looking forward to the new video.
Cam
Sure thing, happy to contribute a bit if I can! Glad that cleared things up. And here’s the video for your reference: https://youtube.com/shorts/VL9cd7AXW9k?feature=share
Going to post a few more soon.