How fast you can move your balls?

Adept Cobra, from Omron, is advertised as fastest Scara robot on the market. I do not know if it's the fastest, but it's fast as hell. Watch it here: Adept Cobra. If it not fun for you to watch, then I don't know what might be.

Maybe in the future I'll make a model of this (or other) small Scara robot - to have model of something fast. I will also make model of giant handlers - to have something big. I'll see what will be more fun.


Best,

K.

3D checked and finished

3D model is finished. I assembled all links together, and measured distances where parts match together - found one place where I was off by 1mm (in 1:1, so negligible on model). I fixed it anyway. I'm quite satisfied with the look of finished model. In total I would estimate it took 10 hours of work, not including writing posts on blog, which took another 5 or so.

Is it similar enough?

Next time - first flattening.


Best,

K.

Model ready

3D model in Sketchup is finished. It is still not perfect, and it has to be checked and corrected where necessary, but it's ready. In next step I will put all links together, and measure all distances between links - to confirm that after assembling each link separately, you can still join them together. Anyway, I'm pretty happy at this point - I was not sure if I can finish this task, and if I will find time to write something about it on a blog - but it seems that it's not so hard if I'll have the time - which is always the biggest bottleneck.


Finished links:

Best,

K.

Aliens, anyone?

If you are interested in paper craft and paper models, you have to check http://aliens.humlak.cz/. Site is old, and not updated, but still contains great models from Jan Rukr. Or check the new page from Jan, which seems to be up to date - http://www.alienspapercraft.com/. Lot of fun models, some hard, and some harder, but they seem to be of quite good quality. Unfortunately I haven't assembled any of them, but they are for sure on my wish list.

Go visit, leave a nice comment, and assemble some good models.


Best,

K.

4/2 != 2

Small progress, but still something. link 4 almost done, - just copy, rotate and flange to connect with link 3. But it's not finished yet, so small update. I skipped part of geometry, which i think will be hard to flatten, and even harder to assemble. It might happen that this part (area where link 5 rotates) will have to be changed, but that will be known after link 5 is completed.

Best,

K.

Links, 2, 3...

Links 2 and 3 are ready. Link two was tricky due to change of thickness - sides are wider on bottom, and due to additional part on top - it has to bend accordingly, otherwise there will be mismatch in assembly. Link 3 was even harder to get proper geometry - but in the end it turned out nicely. I had to drop two small cutouts, as on model they would be 1,5mm x 2mm, which way to small to assemble properly (at least I do not like such small cutouts). Overall I'm pretty happy with the look of model and with the progress which has been made. I expected a lot more problems with keeping up to the schedule, especially since now I have quite a lot of work.

Images of all pieces done, and assembled with .step model for comparison. I think it is similar enough to recognize that this model is IRB120 :)


Something to listen.


Best,

K.

Highlight - Boston Dynamics

If you're interested in robots - and I'm sure you are, if you're reading this blog, then you have to know what Boston Dynamics is doing. I had no idea that there is company making fun robots - not the ones doing hard work, but the ones that can do fun stuff. Check out Spot robot - four legged robot, running on batteries, which can walk. Seems like pretty mundane task, perfected by nature through millenia of evolution, which is hard to replicate with robots - and somehow these guys can get it done. See for yourself:

Best,

K.

Link one done, not much else...

Extremely busy days result in not much work done. Fortunately, still some progress can be made - first link of IRB120 was finished. Small parts are the result, but I think it will be possible to assemble - test run will show it. Next links are planned, and starting from link 4 it would be faster, and less parts as result.

Anyway, this is first link, hopefully all relevant geometry is kept - remember, 70mm.model, so pieces are around 1-2mm in smallest areas.

Best,

K.

All your base...

First small update. IRB120 started, software operational, all systems ok. Generally, I'll stick to 1:10 for now, which is super tiny for small robots - we'll see when we'll be at large handlers. For now, due to extremely limited time (I have max one hour per day) I have to find time to make models, write posts, prepare images etc., so work will be progressing slowly. But that's also ok, if I do not stop - plan is to spend at least half an hour every day to work on blog.

What have I done - base of IRB120. Basically 1/7th of model. Currently at 5 part, only 2 requiring assembly, 3 have to be cut and glued into place. I have no idea how it will look all white, but I hope it will be ok.

Here you can see what I've done. On left .step model of robot base, on right Sketchup model of model. It might not look impressive, but in total it is 20x20x50 millimeters, so it's tiny, tiny. It took me about an hour or so, mostly to get measurements right, and I had to redo part of it due to stupid mistake when making groups in Sketchup. Basically, each part is separate group, which automatically clones touching faces between two parts, which will come in handy during assembly. I will try to finish axis 1 and 2 until Monday, but we'll see if I will find required time.


Best,

K.

Design considerations

Today on Wednesday I'll explain couple of other things which I have to take into consideration with my projects. Not all of them are critical and have to be addressed now, but I'll try to explain them anyway.

1. Scale. 3D models are made 1:1 scale, already simplified. That way I am still able to measure and check critical dimensions, as they should be more or less the same on my 3D and on reference model, but having them already simplified will make whole process faster. I still haven't decided when I will scale them down - probably during sheets assembly, so all coloring will be done with full sized drawings. Target scale for models will be between 1:8 and 1:10 - I would rather do 1:8 to have bigger models and less details, so they will be actually doable, but if scaling is done as last step, I can always change it later on. Target is to have decent sized models of small robots (small robots have ~80cm, so 1:10 will give around 8cm model), and manageable sizes of big stuff (big ones are 4m in size, which makes massive 40cm model).

2. Small brothers. I decided that additionally to 1:10 robot, each one will be also available in simpler form, in 1:25 scale. That way small model can be given away, or kept in the office without much hassle.

3. Test assembly. For each model, before coloring step, I will make test assembly to check if dimensions are right - if something folds up nicely on screen, it doesn't mean it will fold up with real paper. Test assembly will be done without color, using other materials (especially glue) than normal models - test process doesn't have to be pretty, but it has to be fast.

4. Coloring. Will be done at the end, and I do not expect much problems - robots are generally mono colored anyway. Coloring at the end will allow me to check on test assembly if there are some areas where color changes have to be preserved across parts.

5. First model. Decided to make ABB's IRB 120 - smallest one. For couple of reasons actually - it will be simpler, faster and easier to make, and for sure I will mess something up - then it will be easier to rework in the future. Or maybe not, we'll see. Anyway, I do not know how long it will take me to make this model - right now I'm pretty busy, but in the end I want to get it done fast - it's always easy to start project, it's always hard to finish it properly. And .step files are available from ABB website. Second one - Mitsubishi RV-4, also tiny one - probably the smallest one I can make in 1:10 scale, with 250mm arms, so 25mm on model.

6. Upload. All models will be available on this blog in .pdf format, and I also plan to start github site just for finished models - it will be easier to find them. Of course, I know that there are probably less than five people interested in robots papercraft, but who knows - maybe someone will find it useful. For sure I will find it fun.


Best,

K.

Software for crazy projects

What software I'm using, to make a paper model:

1. Getting resources. Since I'll be making industrial robots paper models, it's quite easy - the best option is to get 3D drawing from manufacturer, in some sensible format - usually .step files are available to download. If not, then lot of photos where scale is kept or even measuring real life example is required. Since I'm able to get .step files, I will use FreeCAD (I'll try to put links to all software at the bottom of relevant posts), which looks like very nice tool, and it has measurements available, which is required. Other than that we do not need any features, so I'll not cover them.

2. Drawing in 3D. To make paper model, first I need to create 3D model. I decided to use SketchUp 8, as this is the version I'm familiar with. I know that I have FreeCAD available, but I do not want to start learning another program, which is much more complicated that it needs to be. Sketchup seems to have just enough features for model making, and is really easy to use, so I'll stick with that.

3. Flattening. After creating 3D model of a model, I need to flatten it, as it has to be printed out. Several tools are available, out of which I choose Flattery plugin for Sketchup. It has some quirks, but in general I found it simplest and fastest tool, even though it sometimes requires a lot of clicking, as it cannot unfold shapes by itself. Additionally it allows for re-joining of faces, which is quite handy.

4. Coloring and assembling sheets - this has not been decided yet. Gimp or Inkscape are my two first options - I've heard a lot of good things about Inkscape, and I'm familiar with Gimp, so I'll have to choose one. But this is a decision which has to be done later, so I don't worry about this very much.

Just as a test, and as an excuse to put some images into post, I tried to make simple model in Sketchup to see how it goes, and to remember how I should use it - I haven't done anything serious in couple of years now. For added fun - 5 minutes time limit on drawing. Ready, set, go.


Here's the result:

Not bad for a 5 minute run on midnight. Anyway, nothing more today, so this has to be enough. For Friday small update planned - should have basics of model outlined, at least that is the target.

Best,

K.

Update schedule

It's not the first time I start a blog. I hope it's the first time that I will keep updating regularly, for longer that two weeks. My experience tells me that having schedule will help me, as it will require me to put stuff up - my personal target is to have 2 weeks of content prepared up front, so in case of disaster I can still catch up. So, current schedule:

Monday - major report from work currently in progress - design, test assembly, finishing or upload of model

Wednesday - highlights of fun models - not necessarily industrial stuff, but I'll try to keep it far from military stuff as I can. Or robot stuff - there's so much fun and amazing robots out there, it's worth sharing. If not highlight, then I'll post anything which I find interesting - my be you'll like it too.

Friday - small update of work in progress, or any other ranting related to my projects - no idea of I can make anything done during week.

Depending on how it works out, I'll adjust it as needed. Anyway, this means that tomorrow first large post of work in progress should be done - but I haven't started anything:) In that case I will list software which I'm using - maybe someone will find it interesting.

Beginnings

Couple of years ago idea popped to my mind - maybe I could create paper model from scratch - create 3D model, simplify it, flatten, color it properly, lay out sheets, all in hope that it can be turned into model. At that time I was doing some drawings in SketchUp - seemed that it would be suitable for this task, and easy enough that I would not have to spend two years learning it.

At the same time I switched my work - instead of working with machines from 70s, I started in highly automated enviroment with up to date technology. My jaw dropped when I first saw fully automatic line, with huge (that's what I thought at the time) robots handling parts without human supervision.

Somehow these two topics clicked and connected - maybe I could make paper model of industrial robot? The biggest problem was that I did not had any source material, and making models from pictures only is possible, but is doomed to be innacurrate. But this was years ago. Now I have what I need, so the only thing stopping me is myself.

Of course, I have no idea how to make such model, and I'm not the best model maker either. So it will be very challenging, yet I think it can be done and for sure I'll have a lot of fun making it.