What's new
What's new

Introduction / Cobot machine tending how-to video recommendations?

Joined
Feb 22, 2024
Location
Murica - Idaho
Hey there! I am joining this forum to help turn machine shops into automation pros.

TLDR: I have been obsessed with manufacturing and automation for my entire life and want to help machine shops get automated. I am now part owner of a company that is Universal Robots' top certified systems integrator for machine tending. What how-to videos should I make first to help people who are integrating their own robots?

My background
Early
When I was 4 years old, I told my parents I wanted to be an inventor and asked them what I should go to college for. Being one was a fly fishing guide and the other had a marketing background, they said I have no idea, a mechanical engineer? and from that moment on I was obsessed with manufacturing and automation. In 4th grade, I received a Lego Mindstorms kit and became obsessed with robots.
Highschool
At 14, I finished all of the welding projects they would give me and asked them what the machine in the corner was. The instructor said it was a CNC mill and that he had no idea how to run it, but here is the manual, knock yourself out, so I learned the CAM program that came with it and made my first CNC part before I even knew what a knee mill was. Thanks to the First Robotics program, later in high school, I learned how to write code to program robots and run manual machining equipment.
College
I graduated from Montana State University with two undergraduate engineering degrees, where I learned how to run Haas CNC mills and lathes from who I still believe is the best machining instructor of all time. I had a short stint designing ski/snowboard equipment for K2 as an intern before I realized I couldn't live in the city (due to my obsession with all things outdoors) when I moved back to Montana.
Engineer at a plastic injection molder
My first career job was at a plastic injection molder, where I designed the parts and injection molds, programmed the injection machines, and designed and installed the automation to unload the machines. The company had facilities in Montana, China, and Juarez, Mexico (Sí, hablo español), so I am very grateful for the experiences I was provided and all the things I learned. It was here that I had my most interesting automation experience. They had a robot in the corner of the facility in Montana and no robots on any of the machines. I spent my spare time programming it to unload the parts from the machine after hours. The operators were all pissed off because they thought they were going to lose their jobs, but after it was up and running, they were the first to come up and thank me because they didn't lose their jobs but instead, their jobs became more meaning full as they were upskilled to doing part inspection and other jobs that made their time there more enjoyable.
Haas salesmen
While working at the Injection molder I was at a school board meeting for my grade school supporting my mother who was on the board when i struck up a conversation with the gentlemen next to me. He said he sold Haas machines, and I about jumped out of my chair. I asked him to breakfast and told him I wanted his job. He told me about all sorts of shops i could work at but i told him i didn't want to work at a shop, I wanted his job. Two years after bugging him, he offered it to me. I spent five years selling Haas, Star, Sodick, Zeiss, Matusura, Kitamura, Ganesh, almost every tooling line available, and many other lines across Idaho and Montana. While there, I begged the owner to carry UR robots until he finally gave in, and they took off like wildfire. It was different than most machine sales positions because these shops don't have the resources that folks in more highly populated do. If I wanted to sell anything, I had to be the rigger, I had to show them how to integrate the robot, I had to show them how to run the control, etc., so I would get to know their families, their entire financial situation, and I would help them with business decisions. Because no one else was willing to be this involved, I had about an 80% market share during my time there. The stories of the shops I met in the rural parts of Montana would widen your eyes, and there isn't enough beer to tell you all of them. Of course, there was also a fair share of high-end shops that didn't need the aforementioned that I would sell the larger equipment to, winning through shear effort (once even shaving off my mustache to sell a Fastems system after winning a poker game with a customer(bad idea)). Here, I ran into the owner of Rapid Design Solutions (RDS), and I was absolutely blown away by what they were doing for our customers. When the Haas factory outlet was sold to a larger dealer, although they were a fantastic company, I knew I wasn't cut out to work for a big company, so I called the owner of RDS and asked if I could buy in.

History of Rapid Design Solutions & Lights Out Manufacturing
The founder started on the ground floor of a machine shop, where he eventually became manager. He eventually went out on his own to design all of the hardware firmware and software for a 3-axis gantry robot with a tool changer to test the calibration of Microsoft's touchscreens. He couldn't get the machine shops in the area to make his parts fast enough, so he bought a machine, and when he didn't want to load the parts himself, he programmed an Espson 6-axis robot to load the parts for him. The machine shop across town thought it was cool and asked if he could do the same for them. And then another machine shop wanted the same thing, and another, and another. Eventually, he was installing robots of all different brands for various applications all over the Pacific Northwest. One of the shops asked him to use a Universal Robot, which he was initially hesitant to use, being a fan of traditional robots. Still, he quickly fell in love due to their many advantages over other robot brands. While sitting in a hot spring in rural Idaho with him on a dirt biking trip, I hatched this wild plan to join him and rebrand his company as Lights Out Mfg with this goofy slogan: "What do you do with the lights out?" where we would focus exclusively on creating automation systems for machine tending and expand outside of the PNW to the rest of the country. We have no interest in being just another machine-tending automation brand where users can't modify their systems or do not work for job shops with high-mix work. We have a plan to build a direct part loading system that will run all night where every part is different, every part comes out completed, and no custom tooling has to be made. We aren't there yet, but we are getting dangerously close. Many new products must be created to make this dream a reality, so we are tackling one at a time. We already have systems that have done 4 ops where you only need about 25-50 parts to justify the setup time of the robot that job shops love. Sometime this year, we will release a video of a robot directly part-loading at least 25 different first-op parts into a five-axis without human intervention. Then, we will be looking to add the second operation capability to this system.

My question
We want to empower every shop in the country to use automation to solve the labor problem and bring manufacturing back to the US. It sickens me that China buys more robots per manufacturing employee than America, and I plan to change this. I am going to start a video series on how to integrate Universal Robots for machine tending. My question for you is, where should I start? What videos do you want to see first? Should I start by outlining the differences between all the robot stands, grippers, and accessories on the market? Or should I dive right into technical tips and tricks like, here is how to make one robot waypoint based on another so that when your vise moves, you only have to update one waypoint and not the entire program?
 
We want to empower every shop in the country to use automation to solve the labor problem and bring manufacturing back to the US. It sickens me that China buys more robots per manufacturing employee than America, and I plan to change this.
If this is truly what you want to accomplish, and this isn't just a very long winded way to sell more UR robots, then I think you should make videos covering all aspects of automation, especially how it pertains to small shops. Like, how valuable an auto saw can be. Or how just putting auto doors on machines can save time and effort. In-process probing to do all kinds of things, macros, manual pallets, etc. Spindle grippers have been coming up on this forum more and more, and is how I plan on getting mostly automated if I can get it to work with our parts. Robot arms are nowhere in my plan.
 
You asking those of us that are not fluent - where to start - seems out of place.

If you think that you are the guru (and you very well may be) then you should be able to figger out where to start.


-------------

I am Ox and I approve this post!
 
You asking those of us that are not fluent - where to start - seems out of place.

If you think that you are the guru (and you very well may be) then you should be able to figger out where to start.


-------------

I am Ox and I approve this post!
Thanks for your input Ox. I thought I would see if there was a particular area people were struggling with or wanted to know first before I started on the list I had planned.
 
If this is truly what you want to accomplish, and this isn't just a very long winded way to sell more UR robots, then I think you should make videos covering all aspects of automation, especially how it pertains to small shops. Like, how valuable an auto saw can be. Or how just putting auto doors on machines can save time and effort. In-process probing to do all kinds of things, macros, manual pallets, etc. Spindle grippers have been coming up on this forum more and more, and is how I plan on getting mostly automated if I can get it to work with our parts. Robot arms are nowhere in my plan.
Thanks Matt, I like where this is going. The reason I was going to focus on robots is because the topics you covered already have some really great videos & manuals made on them; however, starting with basics like chip and coolant management, cold saws, ceramic brushes, rotating tools in reverse to remove chip balls and such sound like a great place to start. I'm just worried videos on manual pallets, spindle grippers, macros, and probing might get a response where everyone says ya, no kidding, we've seen this before. The auto-saw is a great point that is often overlooked. It's amazing how many shops have employees standing around in the morning waiting for their stock to get cut before they can start their day.
 
Take a machine that is not set up for automation, and add a robot. Keep it basic, either loading stock onto an air vise for first op on a mill, or loading and unloading a lathe. I want to see several things-

1. A running tally of the cost. The arm itself, gripper, auto vise if a mill, cylinder for the door, whatever it takes to control the whole contraption, etc.

2. A running tally of the time it takes to implement.

That's start to finish, dollars in one corner of the screen and hours in the other.


After it's running you can talk about chip evacuation and tool management and deburring and probing macros and whatever else. Find a basic Fanuc machine, solid but nothing fancy, and make it happen.
 
Take a machine that is not set up for automation, and add a robot. Keep it basic, either loading stock onto an air vise for first op on a mill, or loading and unloading a lathe. I want to see several things-

1. A running tally of the cost. The arm itself, gripper, auto vise if a mill, cylinder for the door, whatever it takes to control the whole contraption, etc.

2. A running tally of the time it takes to implement.

That's start to finish, dollars in one corner of the screen and hours in the other.


After it's running you can talk about chip evacuation and tool management and deburring and probing macros and whatever else. Find a basic Fanuc machine, solid but nothing fancy, and make it happen.
This sounds great! Thank you for the recommendation and inspiration. This will be my first video.
 
The biggest hurdle, in my opinion, is a salesperson telling you how easy it is to integrate the robot but refusing (or being unable) to tell you how. Then when you speak to an integrator, they want to charge $150k to set up a collaborative robot you already own to load a lathe. Later on, figuring out how to send a cycle start signal without a robot interface has been problematic. We just had the robot push the green button.
 
The biggest hurdle, in my opinion, is a salesperson telling you how easy it is to integrate the robot but refusing (or being unable) to tell you how. Then when you speak to an integrator, they want to charge $150k to set up a collaborative robot you already own to load a lathe. Later on, figuring out how to send a cycle start signal without a robot interface has been problematic. We just had the robot push the green button.
Thanks for chiming in; this is exactly the kind of response I was looking for. Sending a cycle start signal would be a great short how-to video. For future people reading this thread, before I make the video, you get a magnetic disconnect relay (happy to send a link if you like) and put it between one of the robot's digital outputs (I'll go over that in a video too) and wires that go directly to the cycle start button. It sounds goofy, but there are many reasons to do it this way based on machine-wiring schematics and safety risk assessments.

$150k is definitely overkill. We are usually around $110k for a full turn-keywhich includes the $50K for a UR10e. Every system is different because every shop has different needs. Most of it is hardware costs. Integrators either overcharge or don't know what they are doing, so it costs more than needed. Anything less than that, and you will end up with a system that isn't reliable or user-friendly because it will be made of cheap components that cause reliability issues or can't be set up for a new part quickly or not easily moved to another machine.

How many hours would you say it took you to integrate your robot?

Robot salesmen (we buy our robots from dealers as well, we just get a discount for being an integrator because we buy more robots) oversell how easy they are to use. Don't get me wrong, I could teach your grandmother how to program a robot to load a CNC machine in about 20 minutes. The thing they aren't telling you is there is a big difference between programming it to do it once and programming it to do it all night long without an issue. This is where the huge learning curve is. What happens if a chip falls in the vise after you blow it off? What happens if someone puts a part in the infeed crooked? What happens if you lose air pressure? etc. I plan to cover all of this; there is just a lot to cover, so I thought I would start based on requests. The reason we have such great reviews with customers is that we are upfront about all of this and walk customers through all the hurdles so they can be successful after we leave. We just shorten the learning curve so they are running lights out immediately and then have the knowledge to overcome issues for future parts and changes they make to the system. The biggest mistake we see is people use unridgid components ei. extruded aluminum stands, grippers with too low of clamp force, unridgid gripper fingers, inaccurate infeeds, etc. Each one by themselves seems like a small issue but all those add up to be the difference between running all night and not.
 
Sending a cycle start signal would be a great short how-to video. For future people reading this thread, before I make the video, you get a magnetic disconnect relay (happy to send a link if you like) and put it between one of the robot's digital outputs (I'll go over that in a video too) and wires that go directly to the cycle start button. It sounds goofy, but there are many reasons to do it this way based on machine-wiring schematics and safety risk assessments.
@jhearons already solved the problem. He had the robot physically push the cycle start button, which is exactly what we do. It's the correct way.

The future of robot integration in CNC machine tending is non-integration. Keep the robot humanoid, i.e. minimally invasive. A human doesn't use a brain implant hardwired into the machine control to command cycle start, and neither should a cobot.

This is where the huge learning curve is. What happens if a chip falls in the vise after you blow it off? What happens if someone puts a part in the infeed crooked? What happens if you lose air pressure?
Probe every part before every operation. No exceptions.

Spindle grippers have a major advantage here. We turn the coolant on while parts are being loaded, so it's virtually impossible for a chip to get in the way. This is after the jaws have already been blown off with a chip fan.
 
usually around $110k for a full turn-keywhich includes the $50K for a UR10e
Go lower. Use a Epson or similar $15k system (or even a $5k Alibaba special). It doesn't even have to run lights out, as long as it's sitting there keeping the machine fed all day someone can glance at it occasionally.
 
@jhearons already solved the problem. He had the robot physically push the cycle start button, which is exactly what we do. It's the correct way.

The future of robot integration in CNC machine tending is non-integration. Keep the robot humanoid, i.e. minimally invasive. A human doesn't use a brain implant hardwired into the machine control to command cycle start, and neither should a cobot.


Probe every part before every operation. No exceptions.

Spindle grippers have a major advantage here. We turn the coolant on while parts are being loaded, so it's virtually impossible for a chip to get in the way. This is after the jaws have already been blown off with a chip fan.
this is where Optimus from Tesla is gonna be an absolute game changer!
 








 
Back
Top