Where and how do I get started with robotic integration in my fabrication shop

    Good evening

    We are a structural steel fabricator and I have been interested in a robotic solution for the last couple years. When we financed our new facility a python or pcr42 or a vortman machine was in the budget. Unfortunately the oil market has crashed and so have our sales. Investing 800k or more int these machines is simply not an option. There is not enough work to sustain a capital investment of this size. This brings me to where I am now. I find myself competing against others suppliers that are in Edmonton with this equipment and more labour force to chose from.

    I feel there is a gap in the market regarding this technology. In the past cnc plasma tables were expensive like the pythons and similar, but technology has made it that almost any fab shop can find a cnc plasma table solution within there budget. Some are not hi definition or have minimal size and or bells an whistles but they all save a tremendous amount of time.

    I have only started looking into the robotic solutions and am trying to educate myself as to "where the technology is" I can't pioneer this field on my budget, but perhaps I can put together some pieces to the puzzle and if nothing else educate myself. I know that given my northern location and lack of labour and overpriced labour I will need to rely on robotics to stay competive and growing my buisness including cutting and welding.

    Some questions I have are
    - do I need a new robot
    - how do I program a robot ( I have seen some offline solutions)
    - because we do structural steel many parts are "one offs" so there is no chance for repeatability. Will the software be automated enough to make this viable, how long will it take to program 1 beam? and next is the software affordable?

    We have drafting design staff operating solid works and I can purchase dstv conversions. I have also designed many of our specialized fabrication equipment so designing and fabricating the "cell" is exciting and completely within my capability.

    This is just a start to my questions.

    can anyone help me with this or have any advice for me?
    where I can start?

    Thanks in advance!

    Dean
    All Class Fabrication
    http://www.allclassfabrication.ca

    Hello. I'm wondering how you are doing. Did you decide to purchase a robot? If so, which one? What size? Did you use an integrator to help build the cell and get everything working for you?If you decided against the robot, what are you doing instead?

    Just curious.

    Building robotic welding systems.

    Well, you're going to need a good integrator, preferably someone reasonably local. You'll need to work with them to nail down what parts of your process are amenable to automation, and what upstream changes are needed to enable that. For example, I've had customers who wanted robots to drill holes down to 0.030" accuracy in their parts... but the parts they were feeding to the robots were only held to a 0.040" tolerance in many areas. So the robot either has to be hyperintelligent (which gets expensive and unreliable), or the customer needed to tighten the upstream tolerance of their supply chain. And they often balked at that -- they wanted automation to somehow magically make a silk purse out of a sow's ear.

    It's never as simple as bolting down a robot where a human used to be and copying what the human did (well, almost never). It will take process changes upstream and downstream of the automation. There's a lot of cost/benefit and trade study analysis involved to get optimum ROI.

    A good approach is to get your feet wet, first. Find a relatively small, stable, isolated process that's amenable to automation, and try automating that first. It's like eating a whale -- take small bites, one at a time.

    Beware of cheaping out -- like my example above, it's easy to waste your money on automation if you don't automation-optimize your process, or if you go with a low-bid integrator who isn't going to explain all that to you, for fear of scaring you away with sticker shock. You want a full-service integrator with a good reputation, especially one that is accustomed to handling new and different kinds of work. You'll need to build a relationship with this integrator for the long term, for service if nothing else. Also, avoid cheap secondhand robots (and other automation hardware) unless you know what you're doing -- there are great deals to be had, but also plenty of lemons.

    Don't expect fast ROI -- automation tends to pay back over the long term, and have rather high startup costs. This isn't a hard and fast rule, but it's the way to bet, especially if you, or your industry, are new to automation.

    Commit. Your first automation foray may actually fail -- that's one reason to keep it small and simple. Unless you can buy a turnkey system that has already been demonstrated to work in similar processes elsewhere, you're embarking on at least a partial science project. That's nothing to be afraid of, but you need to have it foremost in mind, so as to avoid buyer's remorse. Thing for the long term.

    You absolutely will need to train your personnel in the care&feeding of your automation -- to what skill level, will depend a great deal on the scale of your rollout, the reliability of your automation, and your relationship with your integrator (and how fast they can respond -- if they're on the other side of the country, your emergency response time could be measured in days). An idiot-proof system that can be run by monkeys is more expensive than something that can be kept running by trained, experienced operators and maintainers. "Robots as a service" is a fairly new thing in the market, and shows a lot of potential, but I don't know if it's really there yet.

    Make sure your contract with your integrator is well-thought out. Require full documentation, manuals, BOM, and source code up front -- I once found a customer who had bought an entire set of automation from a low-bid supplier in Italy, and hadn't specified these items -- the user interfaces were all in Italian, with no English option, and what little documentation had been written was thin, poor, and also only in Italian. This situation occurs far too often, in my experience.

    Quote from SkyeFire

    <snip> Make sure your contract with your integrator is well-thought out. Require full documentation, manuals, BOM, and source code up front -- I once found a customer who had bought an entire set of automation from a low-bid supplier in Italy, and hadn't specified these items -- the user interfaces were all in Italian, with no English option, and what little documentation had been written was thin, poor, and also only in Italian. This situation occurs far too often, in my experience.

    +1 for that, I once worked on an Italian supplied (I'm in the UK) tube manipulating machine with drawings in Italian, and what appeared to be a home made PLC with no means of looking at the software for fault finding.

Advertising from our partners