Alex Marcy’s Post

View profile for Alex Marcy, graphic

CEO & Founder @ Corso Systems Building great tools for manufacturing companies to better serve their customers since 2007.

Every so often I see a webinar along the lines of “Implementing OEE in under an hour” This is definitely do-able. IF you already have all your data points mapped out. The time consuming part of every OEE implementation I’ve seen and done is figuring out what data to use, what data is valid, and getting everyone to agree on what is the correct data in the first place. “Is the machine/line/plant running” is usually the easiest of the pieces to determine with consensus. Oddly enough cycle time, and especially design cycle time to compare with actual cycle time has been one of the most challenging pieces. This tracks with my experience though. Many vendors don’t build in cycle time tracking into their equipment because it’s relatively easy to measure manually for proving everything meets spec. Design cycle time is usually lost to the ages. Quality is usually one of the easiest pieces to integrate, with the main question focused on WHEN quality data is collected. Most processes don’t have real time metrics on quality so there is more integration that needs to happen to update OEE for a batch when quality is collected. My approach is simple. 1. Pick a running state to start, worry about automated downtime reasons and more granularity as the next step after implementing OEE. 2. Make an educated guess for cycle time and design cycle time then adjust. 3. You can ALWAYS adjust the OEE calculation once you get quality data, so if you have it great, if not it’s okay to ignore the quality portion to start. Get started, learn how it works for you, then adjust and improve. Isn’t that the whole goal of OEE in the first place? May as well approach implementing OEE in the same way 😎

Bret Young

Founder | Patent Holder | Controls Engineer

10mo

It's like a pyramid. OEE is near the top. All the work is on the bottom. That stuff has to be done first. There's no sexy sales pitch for it because every plant is unique. The OEE people are promising you the world and sell you the piece that goes near the top or at the top. They charge you a lot of money then leave you high and dry because they can't do that hard piece. The solution is get an integrator that can do everything from factory floor to OEE.

Scott Chapman C.E.M.

Automation & Instrumentation Engineer | Transformation & Sustainability Leader | Disruptor & Fractional CEO

10mo

Real-time quality metrics are not that hard to implement IF you're willing to pay for it. OEE software seems funny to me because it's a simple calculation you can run on a local HMI and push to the network for corporate dashboards. Alex Marcy is totally correct on the data. I've seen so many companies measuring the WRONG stuff. Custom calculations that make NO sense. Looks good, though 😁!

Drew Thomas

CEO of Oneiro Technologies🏆Winner "Best Use of Robotics-2024"🤝Turn key, agnostic supplier of industrial automation systems up to $50MM | 25+years of integration experience | AMR/AGV/ASRS/WMS/WES/LMS/Conveyor/PandA…

10mo

Alex Marcy this is so accurate for a production facility. I personally integrated an OEE system at a winery many years ago. The time consuming part is definitely getting good data out of the machine center because they are all unique. Good data also requires determining whether the machine has incoming product available to process and a place to send it going out. I can’t believe OEE software companies are making such claims.

Anthony George

CEO and Chief Engineer

9mo

Well put

Like
Reply
See more comments

To view or add a comment, sign in

Explore topics