Hello everyone,
First of all, Thanks to Chris for single-handedly putting together this gem of a robot. It truly democratizes robotics for the general public!
I've been wondering if any member has used this robot successfully in a factory. I want to put this robot for tending to a CNC Lathe in the factory but alot of experienced people in robotics are telling me that it would'nt work over multiple cycles. The job is just to pick up the part (200gms around) and place it in the collet of the cnc lathe.
I'm confident that AR4 would be able to do this job once programmed (as I saw in one of the videos about AR4 tending to cnc machines) but not sure if it would hold up after multiple repeated cycles. Would really love to hear the experiences of people that have used AR4 or tried to use AR4 (or AR3) for any such process before. Please also mention any other disadvantages of the robot that you have noticed Thanks in advance!
I have had numerous people setup the robot for CNC tending in their shops at home and I have tested the robot running for 24 hour periods checking reliability and accuracy but I probably cant answer your question as far as it being in a factory. Hopefully some others will be able to post on this thread and add some info - I would love to see or hear of some examples let me know if there is any other questions I can help with.
Thanks for the reply and apologies for late response.
Would love to know what were the results of your 24 hour check for accuracy and reliability. Did you see any issue with the repeatability over time?
Also, Do you know of anyone regularly using AR4 for cnc tending for long hours at a stretch (Not limiting the venue to be factories but any place in general) Thanks again, looking forward to your reply.
@pradyumna-bang In my testing I just re-tensioned the belts after a break in period but running over long periods I did not have any issues with repeatability. I have seen several people using it for CNC tending but I'm not sure how many hours at a time they have run the robot.
A few words from a general point of view:
Mechanics experience minimal wear as long as there is grease, or a lubrication film. If bearings and gears are not sealed, then the air needs to be dry and clean. Perhaps as a measure, after spending a day in the factory, if it is black what comes out when blowing your nose, then the mechanics may wear out fast.
But I think that an experienced person would also point out that the quality of the individual machine component is important. Bad material quality, hardening and tolerances of rolling elements and bearing rings are negative for life length calculations, e.g. the L10 value (ISO 281).
One way to see the effect of bearing quality is to compare the effect on the life rating by changing the bearing quality in the formula (or lubrication contamination). For SKF, it would be to compare standard SKF with SKF Explorer. If you search on the Internet for ‘skf bearing catalogue pdf’ you can read examples and make your own calculations. You can also search for ‘Bearing rating life’.
On page 88, Table 1, you find that the design and operational environment have an effect on the order of magnitude 3 or 4. With low quality bearings, you will be in the top of the table, and with low load on the robot arm, you will be in the bottom of the table.
Page 88, Table 1 (SKF 2018):
Some less important notes:
For circuit boards, if they are kept clean and are OEM, they are reliable. Electronics reliability follows an exponential distribution. To make it more reliable, you add redundancy. This is important in safety critical systems, like elevators and airplanes. Then software reliability is critical as well.
Regarding the reliability of stepper motors, I think you can read on 3D printer forums, like reddit.
Also, you have some limit switches that you may want to check the datasheets for reliability figures.
Thanks alot for the detailed explanation, that's very helpful!
