Its hard to put a figure on costs, and ROI’s. As a figure that probably doesn’t mean much with out the full facts our down time went from 16% to 0.6%. That’s downtime vs project hours which run in to the 10’s of thousands over the year, over multiple projects and locations.
But then again, we are throwing away parts that are ‘working’ so machine costs go up, downtime goes down and some locations run into hundreds of thousands a day in operating costs that could be held up.
Initial costs is also hard to put a figure on. You could say our system is as dumb as it gets. No redundant servers, sometimes no internet connection. No fancy MQTT messaging queuing system. If we loose one message, we will get the next one a hour later, a week later etc. But, that’s because the bulk of the work is done at machine PLC level. That’s what does most of the calculations, logging, conversions and averages etc. That PLC is there anyway controlling the machine, so really that’s no extra cost. Once I program the data blocks in the PLC I can copy them over and over. So the only hardware cost besides modems and routers is a MQTT device at machine level sending the messages.
Indeed the return is not always tangible. Building more transparency, visibility , observability , instant response , remote monitoring at central location , more and more operational data helps in doing an analysis in real time or as a postmortem analysis to draw actionable insights and inferences . So the incremental cost of such a solution is mostly that of MQTT modules and hardware I guess ?
We developed our own “Remote Units” they are “SMART” meaning they actually manage the location, so if any problems come up they are taken care by the site unit. This way if for some reason we do drop communication to Ignition all site decisions will still be preformed.
We build this box for $1750.00 (USD) and then add what ever sensors we want for monitoring various points.
This box includes PLC, HMI, Cellular Modem, MQTT Gateway, PS/With battery Backup, Antenna.
I am sure that will a great learning for the O&G community! Obviously they will need to go into more details of the individual components inside the box, such as what PLC, what HMI (is it Ignition Edge?), which modules of MQTT etc . But its a great learning experience for the community as a whole.
We do not use Ignition Edge! This is really because our unit needs to manage the location even if there is no communication with server.
This is why we developed our own remote unit for this.
Right now we just show, record all the tags from each location with scada. We query this data in various ways to show Pie Charts, Trends, Daily Totals, Weekly, Monthly.
We have our own “Cellular Private Network” so any supervisor can connect to any location with his/her phone/laptop and make needed changes.
This works for us as of now, but we make changes to the infrastructure all the time.
Thanks for reply . So it’s your own custom solution based on your needs . Perhaps Ignition edge may be overkill if you have very few sensors to monitor .
A while ago I added some info on this post on how I go about it.
Took a bit of time to set it up with the pi having 3 ip addresses and writing the IP tables and modifying the various code. Now I have the pi image saved it’s simple to create a new device.
It may help someone, as above my system is effective but basic. I have no unmanned stations etc.
Kevin, what other ‘collectives’ exist for Ignition? I’m in manufacturing (i.e., inherited a collection of machines & systems originally delivered without connectivity and/or scada of any sort that I’m gluing together into a cohesive structure, often very much against their will and am curious what else is out there. I didn’t see any links back to the mother-ship from the link you provided.
and am curious what else is out there. I didn’t see any links back to the mother-ship from the link you provided.