Robotics

[FROZEN]

Anyone using them in a Job shop production setting???? Loading, unloading, washing, and measuring??? Found this in a earlier post.

 

http://www.mtbtech.net/blog/2013/10/27/Robotics-as-part-of-an-End-to-End-Process.aspx

[Guest]

Job Shops are great candidates for robots. People just don't think they are because they have a deep seated misunderstanding of how flexible and powerful robots really are.

 

They are not just for long production runs.

 

JM2CFWIW

[YoDoug®]

The real question with Automation is what is your defined goal. You can develop automation systems that are easily adaptable to shorter run job shop type work, but they will also have limitations over full blown high volume production automation. Simple automation can give you that 2-4, or more hours of unattended running of a machine. Likewise it may allow one operator to keep 3-4 or more machines running. However it can add considerable time to setups. It can also be costly. 

[Eric@HorsepowerInc.]

We have actually been looking into this for some time now. After going to IMTS, we are looking at the Universal robots from a cost and no cage perspective.  www.universal-robots.com 

These seem pretty robust and easy to set-up and move machine to machine. Best part is no cage is needed to hinder operator work space and valuable floor space. Cost for their largest unit is less than a small lathe, installed with training.

[YoDoug®]

We have actually been looking into this for some time now. After going to IMTS, we are looking at the Universal robots from a cost and no cage perspective.  www.universal-robots.com

These seem pretty robust and easy to set-up and move machine to machine. Best part is no cage is needed to hinder operator work space and valuable floor space. Cost for their largest unit is less than a small lathe, installed with training.

 

We are a reseller partner with UR. We have a UR in our showroom right now that we are planning to set up on a Hardinge mill for demos. That being said, the biggest things that make it hard to switch from machine to machine are autodoors, spare M codes or robot interface, and automatic workholding. Not a lot of machines come with those options unless you order them that way. Also, be very careful about assuming the robot can be run without safety guarding. If your parts have any mass to them or sharp edges you will not pass an osha inspection of your setup. I would look into a small enclosure like these (  http://www.sicron.com/ ). They can be easily moved with a pallet jack. 

[Eric@HorsepowerInc.]

We are a reseller partner with UR. We have a UR in our showroom right now that we are planning to set up on a Hardinge mill for demos. That being said, the biggest things that make it hard to switch from machine to machine are autodoors, spare M codes or robot interface, and automatic workholding. Not a lot of machines come with those options unless you order them that way. Also, be very careful about assuming the robot can be run without safety guarding. If your parts have any mass to them or sharp edges you will not pass an osha inspection of your setup. I would look into a small enclosure like these (  http://www.sicron.com/ ). They can be easily moved with a pallet jack. 

Yo Doug, thanks for the info.  Yes we are aware of some of these potential issues, that being said, all of these can be addressed. All of our machines have spare codes that can be fired, and yes end arm tooling or sharp edges need to be taken into consideration. That being said, we are mainly looking for these to attend our lathes with smaller parts 1-8 lbs.  How have these units held up in the field? Anything to be aware of if we were to purchase some of these? 

[YoDoug®]

Yo Doug, thanks for the info.  Yes we are aware of some of these potential issues, that being said, all of these can be addressed. All of our machines have spare codes that can be fired, and yes end arm tooling or sharp edges need to be taken into consideration. That being said, we are mainly looking for these to attend our lathes with smaller parts 1-8 lbs.  How have these units held up in the field? Anything to be aware of if we were to purchase some of these? 

 

We had issues with one of the older models that would occasionally loose position of the encoder of one joint on start up. We traced it back to noisy power in the customers facility. The newer UR's have absolute encoders though so it is a non issue now. Our typical MO for machines without robot interface is to tie in the discreet IO from the robot to the machines cycle start button and Estop circuit. We let the robot issue cycle start, then the machine gets ready and issues an Mcode request to the robot for a part. After the robot loads the part it gives the  M code FIN signal to the machine and the machine cuts the part. Then the machine issues a different Mcode to signal to the robot it is finished machining. The robot answers FIN and waits for a few seconds to let the CNC read M30 at the end of the program. Then the Robot can issue cycle start to repeat the process. We also tie the fixturing into the Robots IO so it can control it. On a lathe you would need to have access to chuck open/close or jump into the pedal signals from the robot. This poses a safety risk if your machine can allow chuck open/close during rotation. In that case I would issue an M19 from the CNC and tie the robot IO into the M19 encoder signal so you know it is not moving before you open the chuck. 

[FROZEN]

How about inspection out side the machine???

[YoDoug®]

How about inspection out side the machine???

 

If the volumes are high enough to justify dedicated gauging I would recommend Kurt gauging. Then we would use AutoComp from Caron Eng to feedback to the control for dimension control. One of the hardest parts of full automation is ensuring quality parts are being made. This is where you may need to add steps into your machining process to help control quality. The overall goal of automation is to increase throughput and reduce costs. A lot of people get focused on just removing operators interaction, that's not always possible. Gauging can be done by a human and still realize increased throughput and reduced costs. Likewise part staging is a big factor. Say you have a part that is 6" x 6" x 6", you run a 6 min cycle producing 10 parts per hour. To run for 8 hours unattended you need to have staging for 80 part incoming and outgoing. This need to within reach of the robot. You end up with a long conveyor system or stack-able pallets. Both options add costs and complexity. However a few multiple machine cells that have an operator making hourly part inspection and reloading/removing stock/parts will see lowered costs and increased throughput.

 

Being a machine tool distributor we see a lot of different companies and their automation efforts. Some are "kicking @ss and taking names", they are the fastest growing shops in our area and they are taking work from other shops on a consistent basis. Other shops have one automation cell they bought from a distributor and it runs great, for that one part. Even worse we see some shops that had that one part automation and have lost the contract, the robots sit collecting dust. What is the one difference between these shops? An automation engineer. The shops that have invested to have an automation engineer on staff are those shops that are growing. They can adapt instantly and reuse/redeploy automation investments when things change. There is an old saying, If you fail to plan, your plan is to fail. A strategy to move to automation should be well planned. Sometimes a multi-year implementation. With a proper plan the costs of adding an automation engineer will be a justified investment, without a plan it will just appear to be added cost to the bottom line. 

[Guest]

We just quoted a project like that. A machine, a CMM, a Robot, Integration, Caron Eng. AutoComp and TMAC, programming, set up, tooling etc... for some parts that are similar enough to a significant % of the parts they want to run in it. It will be capable of running 24-6. No setup change over time. Can be run manned or unmanned. : These are not "production" numbers either. 5 of these, 10 of those 6 of the other per week. With the "recipe" being what we do for them on the programming, planning, setup, etc... they will be able to replicate and succeed on their own. It's not cheap, but neither is what they are doing and how they are doing it.

 

 

 

By failing to prepare, you are preparing to fail.

:

[John@WMM]

Our shop has been an integrator of FANUC robots for over 15 years. Way back I can remember having one of our robots do some deburring, and yes I was happy about the robot doing the deburring.  Check it out http://www.wauseonmachine.com/automation/automation.html

[crazy^millman]

Good Article about this subject I came across I thought had good information for those looking at Robotics for their manufacturing needs.

 

http://www.toolingandproduction.com/cms/tap/opens/articleview.php?nid=3&bid=408&aid=1764&[email protected]&lk=1&sr=R&tp=Y

 

 

The car manufacturing group Volkswagen has integrated an industrial robotic arm from the Danish manufacturer Universal Robots into mass production at its engine production plant in Salzgitter, Germany. The lightweight robot UR5 is installed in the cylinder head assembly section, where it is responsible for handling delicate glow plugs. The six-axis robotic arm is able to collaborate directly with people without any guards, contributing significantly towards optimizing ergonomic working conditions.

 

 

Enjoy.