Author Topic: Planning for Kuka purchase  (Read 999 times)

atk

  • Forum User
  • *
  • Posts: 4
    • View Profile
Planning for Kuka purchase
« on: March 15, 2017, 08:07:46 PM »
Hi,

We are in the process of purchasing two of the Kuka KR10 900 SIXX.  We are also waiting to get a pitch from ABB but we are leaning towards the Kuka's.  Our organization is new to this.  We have a lot of experience with CNC machines, laser cutting, 3D printing, etc but robots is new territory.  So a few questions:

1) End effectors
- We are looking to be able to assemble, 3D mill (wood, Styrofoam, plaster), hot wire cut, and possibly weld.  We'd like to build our own end effectors and have about $20,000 set aside.  Would you recommend a tool changer?  Should we install a track system?  We are located in Toronto, Canada, would you know where we could source reliable parts?

2) Software
- Kuka PRC and Python seem to be the go to's.  We use Rhino quite a bit, I definitely need to brush up on grasshopper.  Any other software to look into?

3) Any recommendations on literature to read?  There is an online coursehttps://www.onlinerobotics.com/Curriculum#3 offered close to us that we were interested in.  It's more geared to manufacturing but I thought it might be good to do while we are in the prep stages of this.  Any thoughts?

Thanks, any info is appreciated.

AK

Johannes @ Robots in Architecture

  • Lead Developer
  • Administrator
  • Forum User
  • *****
  • Posts: 581
    • View Profile
    • Robots in Architecture
Re: Planning for Kuka purchase
« Reply #1 on: March 16, 2017, 09:47:03 AM »
Hello AK,

Are you sure that you want to go for the R900? Personally I would tend to go for the R1100, as it has got 20cm more reach and shouldn't be significantly more expensive.
As you have got a proper budget, I would definitely recommend to use a toolchanging system, as it makes it much more convenient to switch between different applications. Also, the Agilus has got conveniently placed internal valves (with internal tubing), so you don't need to hang tubes along the outside of the robot. We've been happy with Schunk tool changers, but there are plenty alternatives around. If you want to move the track simultaneously with the robot, you need to use a KUKA motor and need to expand the cabinet - it's quite expensive to do this afterwards, so at least get the larger cabinet size if you are considering that upgrade. Of course if you simply want to re-position the robot between jobs, you can build something by yourself, and then just have the robot communicate via digital ports or Ethernet (with the right software) - or simply move it by hand.

Regarding software, we are of course tending towards KUKA|prc, but there are other interfaces such as HAL that also support KUKA robots through Grasshopper. KUKA themselves offer KUKA SimPro, but in our experience it is more suited towards "traditional" robotic applications.
Concerning literature I would recommend starting with the KUKA manuals, which are quite thorough, and for inspiration the Rob|Arch books (which members can access digitally for free). If you don't have the manuals, send me an eMail to johannes @ robotsinarchitecture . org
When you purchase the robot, also consider if you need any extra software. There are techpackages for load data determination, Ethernet communication, command streaming from KUKA|prc (mxAutomation)... etc. Be sure that you get an Agilus with an X11 connection, unless you want to use a specific safety bus system.

Best,
Johannes


atk

  • Forum User
  • *
  • Posts: 4
    • View Profile
Re: Planning for Kuka purchase
« Reply #2 on: March 16, 2017, 06:59:16 PM »
Ok thanks for the info, much appreciated.  Any preference with Kuka vs ABB?  What are your thoughts on used robots (ebay)? 

Johannes @ Robots in Architecture

  • Lead Developer
  • Administrator
  • Forum User
  • *****
  • Posts: 581
    • View Profile
    • Robots in Architecture
Re: Planning for Kuka purchase
« Reply #3 on: March 16, 2017, 07:17:20 PM »
Hello,

Mechanically I wouldn't see a significant difference, though maybe ABB has less options in the small-robot class. It mostly depends on the local representative and the local support, so while we may be thrilled how supportive and friendly KUKA is in our parts of Germany and Austria, it may be very different in your region.
We try to be rather neutral in these decisions and also work with e.g. ABB and Staeubli for the conferences - it's just that we know most about KUKA robots and therefore use them in our research.
I haven't ever heard any complete nightmare stories about used robots from our users, so that may show that the robots are really quite robust. That being said, there is always a risk, of course. We've dealt with Global Robots in the UK before and I know that they have shipped to the US - but we cannot vouch for an external company. Also, there are hardly any small, used robots, if that is still what you want to buy.

Best,
Johannes

atk

  • Forum User
  • *
  • Posts: 4
    • View Profile
Re: Planning for Kuka purchase
« Reply #4 on: March 27, 2017, 10:15:32 PM »
Hi Johannes,

We are now looking at some used robots.  We found some local resellers and are waiting on quotes.  Most of the used machines are quite a bit bigger and so now my concern is space.  The room we have set aside for the robot arm is 5.5 m x 4.7 m and 2.5 m in height.  How much space do you usually give a robot arm?  I know that's vague, but is there a standard area they recommend based on the size of the robot?

Thanks,

AK

Johannes @ Robots in Architecture

  • Lead Developer
  • Administrator
  • Forum User
  • *****
  • Posts: 581
    • View Profile
    • Robots in Architecture
Re: Planning for Kuka purchase
« Reply #5 on: March 28, 2017, 09:26:58 AM »
Hello,

Well, a KR150 - which is probably one of the most common models - has got 2.7m reach fully stretched, so 5.5m are plenty and 4.7 still enough, considering that you generally don't use the robot fully stretched. The height is not much, but should be sufficient, considering these robots are slightly over 2m in their base position. It can stretch to 3.1m, though - so you will either have to be careful, or reduce the workspace of the machine.

An issue to consider is safety: It's convenient to be in the same room as the robot, but outside its reach. So having slightly more room wouldn't be bad, though you can still limit the robot's axis range to get a bit more "save" space.
If you don't need to work all around the robot (note that its base can rotate +/-180, not the full 360) then it probably makes sense to place it 1.5m or so away from the back wall. With the 2.7m of the robot, it would take up 4.2m, so that 1.3m are still left to install a fence, or at least mark the space where the machine cannot reach. Or consider adding a window into the wall, so that you can observe the machine safely. Of course all of that is not official safety advice ;)

KUKA provides convenient 2D drawings of robots in their download section (look into the Archive for older machines) that you can place in your plan, and you can of course also use the downloadable version of KUKA|prc to check reachability.

Best,
Johannes


atk

  • Forum User
  • *
  • Posts: 4
    • View Profile
Re: Planning for Kuka purchase
« Reply #6 on: March 29, 2017, 10:54:10 PM »
Hi Johannes,

We are looking at these two models:
http://www.globalrobots.com/product.aspx?product=22837
http://www.globalrobots.com/product.aspx?product=24880

What are your thoughts on either of these?

We are also looking at purchasing this system (minus the robot arm):
http://www.robotics.ca/wp/portfolio/milling-sculpting/

Will the Robots in Architecture software be able to interact with this system (robot and the turntable) or would that be with python or arduino?

Best,

AK

Johannes @ Robots in Architecture

  • Lead Developer
  • Administrator
  • Forum User
  • *****
  • Posts: 581
    • View Profile
    • Robots in Architecture
Re: Planning for Kuka purchase
« Reply #7 on: March 30, 2017, 07:36:04 AM »
Hello,

Both robots look fine, from the description there does not seem much of a difference except the payload (150kg vs. 210kg). Having a KRC2 ed05 controller is definitely a good idea for convenience such as working USB ports.
If you use a KUKA robot with a turntable (in a synchronized way, that is) you need to use a KUKA motor anyway. So assuming that, it should work fine through KUKA|prc - we may need to implement the 3D model of the turntable, but that's not a big problem. Global has got a cheap turntable as well where they basically recycle old robot motors. I doubt that it can support 4 tons like the one you linked to, though.
Don't forget that you usually need to put in additional hardware into the control cabinet to drive a seventh axis. So if you buy a used robot, ask the seller if they have the right hardware in stock and can already install it.

Have you asked the guys at New Age for a quote on the system? Because I would assume the accessories and software to cost at least as much as the included robot itself, as MasterCAM already costs half a robot. That being said, it looks like a really nice setup!

If you are planning to do mostly milling, be aware that you may run into problems with the file size. There is no rule of thump how large you can go, but be aware of the issue. For KRC2 there is a KUKA techpackage called KUKA CAMRob which allows you to read positions off CSV files with extremely large sizes - that's supported through KUKA|prc as well. Our component for importing G-code from e.g. Fusion 360 also supports reducing toolpath complexity, if necessary.

Best,
Johannes