Input requested on your IIoT experiences and applications

Pretty vague topic title, I know.

We're currently upgrading our infrastructure a bit to support a new software system by Memex (Merlin). Our intent is to modernize and increase shop-to-office communication more than we can by simply walking around or talking to leads who honestly have more productive things to do than brief us. Our goals are to monitor machines and collect data on downtime codes so that we can supply attack the real enemies.

For one example that's been a stick in my craw is our lack of an actual tool crib or tool crib attendant. Our tooling area is open and we track usage of inserts and common tools, but every machinist sets up his own tooling, or the lead sets it up for him before the job is run. IMO that should be done offline by someone to keep the "setup time" for each job to a minimum. All offsets should be recorded and ready for the machinist to enter. I think that once I collect and analyze the 'down time code' for "tooling setup" I can quickly show management the payoff for a tool crib attendant that can do the work for them.

I would like to know if anyone has implemented some form of "IIoT" approach to their shop, what you did, what you regret and what you love. To be frank, I had nothing to do with the decision to take this approach, it came from the highest level and most of us only found out when we were told "this is happening" so I'm trying to do my best to make it as effective as it can be, and learn what I can, to contribute positive and successful results.

I've found some good practical real-world literature but for the most part, I get a bunch of bullshit that raises all kinds of red flags. It's like 50% of the literature I find is buzz words, acronym soups, and vague crap just 'selling' the idea, or trying to say what a REVOLUTION it is, rather than telling you any details.

Merlin is good stuff..... we are also implementing that. Have talked to several manufacturers that utilize it and it is good and it does help track down where the issues lie. Mazak in Ky uses it and it reduced downtime significantly after it gave them the information on where the problems "really" were......

I had a custom system for a while as an experiment during the early 2000's. It was citrix based and running a custom application. It was limited to only telling when the machine was down and up, requiring operator input (prompted) to determine WHY. (without the WHY, it is useless). Merlin has the capability of getting the faults out of the machine so at least you have that to go off of.

I used that test system to knock out about 15 hrs of downtime per week (24/5 operation) just in little things that were causing 1-5 minute stoppages on automation production equipment. Little hang-ups that nobody even noticed or paid attention to because they had been dealing with it so long it was just SOP. Those 1-5 minute stoppages were amounting to 20+ hrs a week in unexplained downtime.

Thanks.

The downtime reasons are my biggest eagerness. That's what seems to be the most frequently touted benefit. I believe we do have a great chance to improve in that area. I already know a few things that will most assuredly be backed up by hard data. However, my observations and suggestions contrast what some machinists believe the 'real problems' are. Sometimes right, sometimes wrong, but I believe most of the times we disagree it is because they see a 30 minute task they enjoy or don't mind doing as less of a problem than a 5 minute task they dread. Or they don't realize it took that long because they were in the groove or don't realize there is "a better way".

Do you know of anyone who's adapted the software to monitor tool life or any other "predictive maintenance" aspects?

I'm intrigued.

This Memex Merlin system looks neat, I've never even heard of it before. My first thought is that it's going to sticker shock me right out of pricing, but just from reading though their webpage, it looks like it is suitable for integration with my ERP (ECi M1, uses SQL databases, .NET framework and also Crystal Reports) - and solves many of the shortcomings that my ERP has regarding work flow management.

Ballpark figures for getting this system implemented? ... $10k, $100k? PM me if you'd rather not say.

One of my big problems right now is not getting enough useful information as feedback from my machine shop about job scheduling, WIP progress and a comprehensive maintenance record/plan. I know my ERP can be customized to satisfy this, but I am expecting that it will be in the $25-$50k price range to execute that. If there is a turnkey package that can talk to M1 and provide me with USEFUL information, not just raw data.. that would be worth looking into, especially if it is at a comparable price point.

No clue on price, sorry.

I don't know what our scheduling plans are for this system - that's more of our plant manager's wheelhouse. I might ask him later. Right now, our current SOP is to have our guys 'batch' their time into the time clock system at the end of the day rather than clock in/off a job as they start/finish. It makes it easy to divvy up time when you're running 2-4 different jobs, eliminates problems of "shit, I just got pulled aside for 30 mins on this job, now I gotta tell HR to fix my time clock", and other such 'job shop' issues. So we basically don't know who is working on what, or which machine is running what, unless we put eyes on the floor, which isn't difficult.

However, since Merlin is more about the /machine/ clocking in/off a job, rather than a person logging labor hours, I believe it will be much more useful for real-time scheduling. It seems to separates man-hours from machine-hours which we previously have not done.

Sorry I can't help on price. I typically keep my nose out of the dollars unless I'm begging for a budget or purchase req approval

Bumping to the top for the new week..