Best practice: Parametric placement with automatic path generation -> Software ?

    Hello,

    I am currently searching for a way to automate the proces of "parametric placement" through a software suite.
    More precisely: our company is looking for a way to automatically gnerate a program (paths & logic) for a given parameterized CAD assembly. The idea is to mark cartain center- (or reference-) points of parts of an assembly with an attribute that a RSS software can interpret as a placement point. The software should then create a program that successively places given parts at the respective location until all positions are equipped. The main goal is to keep this process free from human correction in most cases.

    I personally did take a look into RoboDK and I am about to also test Kuka.Sim, because we are already customers at Kuka.
    Most of my experience is based on work with a KR22 and KR 30 HA, but I only have online programming experience. So there was one task and one program to work through the task which is running during production.

    Now we want to be able to cover a huge variety of products (only with simply placing tasks) and also make it possible to easaly create, simulate und deploy the respective program in the following way:
    Import of parametric assembly -> automatic creation of all needed programs -> simulation/ collision detection & check by human -> creation of robot program -> deployment and execution of program.

    From what I saw RoboDK does 90% of the things I am looking for. Also we are about to get a robot from ABB so I like the fact, that RoboDK makes me independent from the brand. Unfortunately even with the awesome tutorials made by Jeremy and provided by RoboDK via youtube, I wans't able to find a way to create automation scripts (my guess would be the integrated python script capability), e.g. I import a part and automatically look for certain points that are for example given by a csv. These points are then used for path creation.

    I would love to hear ideas or alternative ways to get closer to a dedicated solution. Also if someone could point me to documentation or more tutorials from RoboDK, I would appreciate that. We are a medium-sized company that just recently started with robotic automation, so we are not looking for cutting edge software tools (compared to automotive industries). Are there affordable alternatives next to RoboDK that could fit our needs and that are worth a look ? I am tempted to go with RoboDK but I want to do a solid comparsion first.

    Looking forward to hear from long-term robot experts :smiling_face:
    Thanks in advance ! :smiling_face:

  • kwakisaki January 6, 2025 at 10:08 AM

    Approved the thread.

    It sounds like you’re working on an exciting project with a lot of potential for automation, and RoboDK seems like a great fit given its flexibility with different robot brands. The challenge of automating tasks like path creation from parametric assemblies is definitely an interesting one. If you're looking to dive deeper into automation scripting, checking out RoboDK’s forums and community discussions might offer you more specific examples, especially with Python. Additionally, if your team is working on any academic or research projects related to robotics or automation, seeking <Link removed> could be a valuable resource to help structure and document your work, ensuring everything is organized and clearly communicated. Best of luck with your comparisons and integration process—I’m sure it will be a successful journey!

    While I think Python in RDK is probably your best bet, the main trick for making this work in RDK, KUKA-Sim, or RobotStudio, is going to be having enough data, in a consistent reference frame for each part, to allow the software to "snap" the robot TCP accurately to each part in the grid. I think RDK gives you the most "openness" for creating this path-automation tool yourself. I can't speak to how well RS or K-Sim do this.

    You probably won't get very intelligent collision avoidance or robot config management, either, so you'll probably still have to handle that yourself.

    It appears that you are seeking a solution that completely removes the human brain's cognitive analysis. spatial perception, and problem solving functions from the robot programming and integration task. I think it will be difficult to find a COTS solution for this. OK, I'll be negative: impossible to find. I've been involved with offline programming technology for a long time. Customers or Managers wanted a miracle software solution so that they would not require highly paid robot programming personnel. The solution was never developed.

    The solution would require a lot of background programming that would accommodate a lot of variables and random events that would inevitably occur. Products arrive dimensionally correct, but vary within a tolerance (and sometimes out of tolerance). Actuators, sensors, and bearings degrade over time. Friction and static vary with humidity. 3rd Shift operators tinker with the bits while no one is looking.

    That level of programming is prohibitively expensive for a general-purpose COTS product. The digital model for simulation is built to 16 or 32 (or perhaps it is now 64?...I'm getting old) decimal place accuracy. To make a simulation-developed program work perfectly would require the physical installation match the digital model to a high degree of positional accuracy in 6DOF. This requirement can be attenuated somewhat if the process tolerance requirements are loose. But all bets are off if an inattentive janitor or material handler puts a trash can or parts box in the wrong place.

    The optimistic Engineer in me says that it is possible and can be done, but I doubt anyone would want to pay the price or schedule required to develop such a specialized solution. And this discussion doesn't include the required product support to keep it compatible with the latest CPU or OS updates.

    Given the current state of the art, IMHO your best course of action for success in the short term is to use a suitable simulator to achieve that 70%-90% reduction in programming labor, then use skilled humans to complete the task. And, later, to fix things when they blow up or change.

    Quote from SkyeFire

    While I think Python in RDK is probably your best bet, the main trick for making this work in RDK, KUKA-Sim, or RobotStudio, is going to be having enough data, in a consistent reference frame for each part, to allow the software to "snap" the robot TCP accurately to each part in the grid. I think RDK gives you the most "openness" for creating this path-automation tool yourself. I can't speak to how well RS or K-Sim do this.

    You probably won't get very intelligent collision avoidance or robot config management, either, so you'll probably still have to handle that yourself.

    We actually want to start with a simple pick and place process and the geometric data is very well defined, because we actually have the real part & an automatically created digital copy in a CAD program. We want to check if the geometric discrepancies between the digital version and the real part creates problems for placing precision. Eventually we then (plan) to use mechanical or optical referencing tools to enhance the placing precision. But that is not the problem we want to solve right now.

    The main idea is to create the code automatically e.g. the path is always the same counterclockwise placing of each part one after another with dedicated pick up locations. Only the placing position and number of parts to place varies. Because we have a lot of different geometric variations (e.g. different lengths), we want to have a completely automatic output of these (relatively simple) robotic programs. Maybe there is a need to reteach points on-site, but that is also a different problem.


    I will have a closer look into python scripts in RoboDK. I tested both RoboDK and KukaSim for a little bit now and RoboDK seems to have the tools to test my initial idea. At least creating programs and positions is so easy that I could imagine a simple code that automates the process.

    Quote from TygerDawg

    It appears that you are seeking a solution that completely removes the human brain's cognitive analysis. spatial perception, and problem solving functions from the robot programming and integration task. I think it will be difficult to find a COTS solution for this. OK, I'll be negative: impossible to find. I've been involved with offline programming technology for a long time. Customers or Managers wanted a miracle software solution so that they would not require highly paid robot programming personnel. The solution was never developed.

    The solution would require a lot of background programming that would accommodate a lot of variables and random events that would inevitably occur. Products arrive dimensionally correct, but vary within a tolerance (and sometimes out of tolerance). Actuators, sensors, and bearings degrade over time. Friction and static vary with humidity. 3rd Shift operators tinker with the bits while no one is looking.

    That level of programming is prohibitively expensive for a general-purpose COTS product. The digital model for simulation is built to 16 or 32 (or perhaps it is now 64?...I'm getting old) decimal place accuracy. To make a simulation-developed program work perfectly would require the physical installation match the digital model to a high degree of positional accuracy in 6DOF. This requirement can be attenuated somewhat if the process tolerance requirements are loose. But all bets are off if an inattentive janitor or material handler puts a trash can or parts box in the wrong place.

    The optimistic Engineer in me says that it is possible and can be done, but I doubt anyone would want to pay the price or schedule required to develop such a specialized solution. And this discussion doesn't include the required product support to keep it compatible with the latest CPU or OS updates.

    Given the current state of the art, IMHO your best course of action for success in the short term is to use a suitable simulator to achieve that 70%-90% reduction in programming labor, then use skilled humans to complete the task. And, later, to fix things when they blow up or change.

    Thanks for the answer, I think what you said is pretty much spot on with my little experience in robotics. But we still want to eliminate the need to create the very repetitive programs by hand, even if you only need to copy a program, change the CAD part, change the snap on positions and finally export the program again. If that could be handled by a script that would be awesome. Do you guys maybe have examples for these python scripts in RoboDK ? I don't mind if the application is completely different.

Advertising from our partners