Monday, May 16, 2016

Not of Interest to the Unplugged Shop

A video I recently came across which may be of interest - indeed, I would think thought-provoking - to readers here. One would not normally associate CNC work to irregular timbers and forked branches, and yet:



It is intriguing to glimpse what is coming in terms of wood processing machinery and techniques, but sometimes I am taken unawares by what is here already.

9 comments:

  1. Whoa, now that is awesome! I wonder how they extrapolated the 3D mesh from the tree itself; you'd have to have 360* view to do something like that, right?

    --Josh

    ReplyDelete
    Replies
    1. Josh,

      I imagine they scanned the timber forks with some device and imported it into a program which created the mesh and then extrapolated the centerlines.

      ~C

      Delete
  2. That's a whole lot of time and money spent for a simple roof frame for a barn. Is this Bill Gates' barn? Jeff Bezos?

    ReplyDelete
    Replies
    1. Derek,

      thanks for the comment. If you'll note the title of the video, "London's Architectural Association School launches specialist robotics course", the process shown will make more sense.

      ~C

      Delete
  3. I think they used something like a laser scanner to create a point cloud. And then a tesselation algorithm to create the triangle mesh from that. You could put the scanner on the robot arm as well. It looks like they took the bark off before processing. I wonder if you would get a good dataset if you were to scan a fork with the bark still on; I would suspect not (depending on the species of wood).

    Did you notice the 3-point support? A very clever way to completely immobilize the trunk without overconstraining it.

    The video makes it all look very automatic. But I've got the feeling that there was still a whole lot of human work involved. There might be software for the individual tasks, but making everything work together...

    One of the challenges is that a robot is basically controlled by varying the angle at the different joints. So often a certain point in space can be reached by multiple combinations of angles. So if you're not careful and align two axes your robot could end up in a gimbal lock.

    Before we bought a conventional 5-axis CNC machine, we looked at an industrial robot with a milling spindle attached. The thing was that a robot is not as accurate as a milling machine because of the amount of play in each joint. If you were to make the robot plane a surface larger than the diameter of the tool, you would definitely be able to feel the individual milling paths. OTOH, if cou can live with tolerances of say 0.5 mm, a robot would be fine.

    Last time I checked you need special CAM software suitable for robots. Most CAM programs assume movement along the X,Y,Z,A,B axes, which is rather different from the four to six joints of a robot.

    ReplyDelete
    Replies
    1. Roland,

      some great information there. It makes sense that a multi-joint robot arm would not be able to hold to the same tolerances as a milling machine. Hadn't thought of that. I imagine for that project that tolerances of 0.5mm were acceptable.

      I find the topic of robots and their increasing utilization in all sort of industries to be a fascinating one. I often view the situation in light of my own job security, for one thing.

      ~C

      Delete
    2. I must say the the application above seems somewhat atypical.

      Robots are really good at one thing; blindly making the same movement over and over again with good accuracy. We use a robot to e.g. apply a bead of glue on a carbon fiber part to bond some aluminum profiles. This uses less glue and is more constant and repeatable than doing it by hand.

      But programming a robot can be a significant effort, so you generally don't do that for a single piece (unless you can automate the programming, but that's probably not cheap).

      I say blindly because most industrial robots do not have eyes. They're basically just an arm with some tool attached. So for instance in the milling application above, the robot can in general not detect if a tool is worn or broken. Nor does it know what the tool length is. It just performs the programmed motions.

      They're also quite expensive. And the expense doesn't stop with the robot itself. You'll need a safety cage with a door or light barrier so the robot doesn't punch anybody. And you'll need some kind of fixture for the part, plus detection that a part is actually there. Et cetera.

      With regard to job security, I don't think robotizing your work would not make sense given the small series that you seem to make. High-volume woodworking shops are a different story.

      Delete
    3. Roland,

      thanks for the detailed comment.

      My thinking about robots vis-à-vis my own business is not revolving around a 'keep up with the Jones' sort of thing. I'm not contemplating getting a robot for the work I do.

      My thinking about the topic was more along the lines of "what sort of things can robots make?" and "is the business I do threatened by that"?

      I had been thinking that fitting of irregular logs together was the provenance of scribe fit timber work, something only done by hand, and to see the task accomplished by modern methods was both fascinating and unsettling.

      It may not be the perfect task for a robot to undertake, but I'm not sure it would be a whole lot slower than fitting such timbers by traditional methods, which is also slow.

      And maybe the limitations you speak of (lack of eyes, non-perception of tool wear, etc.), are but temporary constraints for robots. It seems to me that most human operators of cutting tools are similarly clue-less when it comes to detecting or dealing with tool wear in a timely manner, and using tools when they are past dull is by far the norm, not the exception. I'm guilty of it myself at times. Tool wear and replacement are largely economic/scheduling issues, especially given a repeated use in a given material, be it MDF, plywood, or a species of solid wood.

      And I don't think volume necessarily has to be a justification for using such a tool. Wendell Castle, for example, uses a robot arm just like the one in the video for his projects, and they are hardly what you would call production pieces.

      I'm interested in the topic of robots and automation as such technologies seemed poised to eliminate a lot of jobs in the near future, particularly on the production side, and I'd like to be seeing the train coming down the track so I might more easily step aside (or get on board, as the case may be), rather than being steam rolled while looking elsewhere.

      ~C

      Delete
    4. You've got a very good point.

      I hope that there will always be room for craftsmanship; our world would be a poorer place without it.

      Like you've already discovered, a mill is a very handy piece of machinery to have at hand. And a CNC mill (like the robot in the video) even more so. Once we got our CNC mill we kept finding new ways to use it. Now we wonder how we ever did without one!

      Maybe you could take a course in robotics or CNC milling just to get an idea of what it can do?

      Delete

All comments are moderated, so if you're planning to spam this, know now that your clicking and pasting is in vain. I do read the comments before posting, so your mission is doomed from the outset. All this time and effort trying to put your inane spam onto blogs -- is this how you want to spend your time on earth?

Please do me the courtesy of appending your name to your comment, even if posting under the 'anonymous' option. No name = deleted.

Comments NOT accepted include:

-those containing links unrelated to blog content
-spam of any kind, or ham for that matter
-did I mention that attempted spam postings will be non-starters?