Quality sat down with Stefan Friedrich of New Scale Robotics to explain why quality is often one of the later steps to automate. Adopting a robot or cobot can be intimidating but it doesn’t need to be. Here he describes what quality professionals need to know. Hint: it’s not programming.

Podcast Excerpt

Quality: You recently worked with Quality on an article about automation and kind of talked about a few different trends and things going on. So to start off, could you tell us what's new with automation these days?

Stefan: I think generally automation's getting easier to use. More user friendly for people that don't have software experience or any prior robotics experience. And that's really because one of the real forefront technologies, which we use a lot and we help customers integrate, is collaborative automation. So, six access robots that are collaborative, that have built-in safety features and easier to use software. Those are really becoming more widely adopted and available. And it's really encouraging that third party companies that make accessories for those robots easier to use with better software interfaces that kind of more emulate your phone and apps that you use on your phone. So, I think overall, that's the direction the industry continues to move in because it certainly helps with adoption. Particularly for smaller manufacturers that don't necessarily have the depth of expertise that a really large manufacturing site might with a dedicated integration team or robotics team or software team that could deploy the robot pretty quickly using more traditional technology.

Quality: I like in the article you were commenting, you don't have to automate steps one through 10, eight through 10, just pick up part of it and go from there. Does that make sense?

Stefan: It's certainly the new technology is helping people automate more incrementally, which is this idea of incremental automation is just that it's if you have a really big, long manufacturing process, let's call it 10 steps, you don't have to tackle the entire process and automate every aspect of production right away. You can take it a little more iterative approach where, you take something that's really easy and repetitive as part of that process and throw automation at that. And while it doesn't solve the whole problem, if you're trying to automate everything, it at least reduces some of the redundancy and you kind of get a jumpstart on being able to tackle that automation project as a whole. And you build little successes over time by taking things one or two steps at a time. And especially for companies where you don't have the cash to throw a million dollars towards automating the whole project, you could maybe throw 50 or $100,000 and automate one step in that process, get an ROI pretty quickly and then maybe tackle the next project a few months later or the following year. And eventually you will build up to hitting your whole goal, but tackling things a little more, one step at a time, really helps divide that big goal into smaller, more manageable projects and expenses. And one of the big benefits of incremental automation too is you really reduce the risk of the project because you're tackling a smaller portion of it at once. And it also dials in the schedule and a lot of projects suffer from scope creep and broadening as you start the project and tackling things one step at a time can really help manage that risk as well. That definitely makes sense. A huge project, like you think of some people adopting new software or something could be a months-long process and maybe it doesn't even work well at the end.

Quality: Can you tell me a little bit about how the process would work? Like the first steps of adapting a robot or cobot or it's kind of automated system.

Stefan: It depends on where you're trying to apply automation. I can talk about quality as one example, since that's what we focus on and that's what your publication focuses on. So, we at New Scale, we're helping people every day automate dimensional gauging. So, usually we start by picking a part or a family of parts that are moderate volumes. Typically with these newer technologies, we're working with people that have parts maybe in the 10,000 parts a month range or even much less than that. And if you start with kind of one of the higher volume, simpler parts you have, or a family of parts that are quite similar, take a look at where your team is spending the most time gauging that part and try to apply automation to that step first. If someone's sitting there with a micrometer and a height gauge right after an operation on a lathe, try to automate that step first. If you're later taking those parts and applying them to the CMM or other kind of more advanced gauging technology in the QC lab, don't start with that because interfacing to the CMM or another larger expensive piece of equipment is gonna be complicated. Usually you run those at lower volumes. And that's just a bigger, broader scope than automating a few hand tools.

Listen to the rest of the interview below.