I just thought I'd mention it since I didn't already see it mentioned. I work in industrial controls and been using node-red for almost 2 years now. I'am generally alone in this when I talk and work with other control's engineers however I was glad to see that June's edition of "Control" magazine which is a process automation magazine mentions Node-Red as a major IIOT (Industrial Iot) player on pg. 24! Woo!
Another controls engineer here. Agree, not many have heard of it or used it. Can thank OPTO 22 and FreeWave for helping to bring it into industrial automation.
Hey ThatCadGuy,
I'd love to hear your take on the forementioned products and if you have used them. I still am using mostly Allen Bradley, Automation Direct and Mitsubishi PLC's. I can throw in a Node-Red device into a SCADA system everyone once in a while 'if they let me' and usually the selling point is I can do it for almost no money on a micro controller as an add-on. I'am certainly interested to see more core-devices that support these types of systems. I hate factory talk and pretty much every other SCADA system I've had to deal with and yearn for the day that something like NR takes over and I can build real tools and can actually do something without purchasing a million 3rd party add-ons. Also factory talk and other SCADA visually - looks like garbage. It's amusing that open source products show how far ahead er rather behind these other products are.
Hi @netwingduck, so OPTO 22 has their groov EPIC system (https://www.opto22.com/products/groov-epic-system) which is kind of a mid-range PLC system and it comes with Node-RED built-in; they even have their own nodes preinstalled. I have not used this -- too expensive, like $2000 with an I/O card or two. But OPTO 22 has REALLY nice stuff though and they are pretty spot-on with their software vs. most other industrial automation manufacturers. If you wanted to try to ditch PLCs but definitely need PLC functionality, I would go to them. AFAIK you can run Node-RED and their control engine simultaneously.
FreeWave, which comes from the world of industrial radios, has their ZumIQ lineup of radios and standalone boards (https://www.freewave.com/products/zumiq-edge-computer/) -- much more affordable, especially if you need hazardous location approvals like I do. These are more of just 'edge' computers though, no I/O, although they have universal I/O products that integrate with ZumIQ. Their standalone ZumIQ board is around $400 or so. Their stuff is pretty good but they are still developing the ZumIQ line -- some bumps, such as proper RS485 support for all use cases of their serial ports. If you are interested I might be able to get them to hook you up with a sample, especially if you can help them do a case study. I did one with them and we got a few radios and ZumIQ boards for free. They're a cool bunch.
MOXA has a lot of edge computers. I have a UC-8100 sitting on my desk at home that I have not messed with yet. Their Linux offering seems very mature and refined. I had to contact support with some questions about how they had things set up and they were pretty quick to respond.
I have also looked at Beckhoff's embedded PCs, but a bit expensive and their website is really lacking in terms of helping people new to their products get a system together -- they have a ton of stuff and it's not clear what fits on what.
on what you said about other SCADA/PLCs. 100% with you. Node-RED has a lot of potential to overtake a lot of the outdated and/or crappy stuff out there if the stigma surrounding non-PLCs can be lifted. We are under a lot of pressure to price things extremely competitively so we have been doing micro-PLCs like the MicroLogix 1400 and IDEC's Modicon re-brands for a while. Definitely don't have room to do a big ControlLogix rack on our piping systems (I'm in oil & gas). We can get execution times on par with the MicroLogix using Node-RED on an ARM computer.
Glad you responded -- I was actually thinking about coming back to this to see how you were using Node-RED in your industry. I am using it as a flow computer basically -- have a lot of math to do and using Javascript in function nodes is a lot nicer than in ladder logic!
Cheers! Sorry for the super-long response!
Thank you for responding, so refreshing to hear from another controls engineer that uses NR. I do OEM stuff for manufacturing and we make industrial furnaces. I come from a SW engineering background and I kind of landed in industrial because of the SCADA iot stuff and my networking background. I'm no EE but I've been scrapping along the past 2 years catching up and getting the ladder stuff more solidified. What we needed NR for was partially to bypass IT departments and for safety/proof of process. When we build a furnace we have a final stage where it has to be cured and this often times happens on-site at the customers plant. I originally had a version that I tried to interface with over LAN and Wifi but the politics with IT departments turned that into a total nightmare and made it impractical. Shifting the liability over to us, it actually makes it safer and the process more streamlined to have one of our techs in contact with the furnace during the curing process so he doesn't have to 'camp out' the entire time during this 3-4 day process. If something goes wrong at 3am or whatever random time then he gets alarmed. I had done research for equivlent SCADA systems using kepserver and datahub and they were just astronomical AND they didn't want to give us write control or charge us per tag. I'm not doing anything fancy, math-wise the main thing I'm doing is logging to sql, texting alarms over SMS (twilio) and running a historian with the graph. I setup my own open VPN server in the cloud and am using an industrial 4g router to connect to a BeagleBoneBlack that I put inside an industrial enclosure with all industrial components. I connect the device using all rugged stuff like M12 connectors and a NEMA 4 enclosure. Its actually pretty funny to have this 80$ micro controller surrounded by all this stuff but its in really nasty places like foundries. They have been working like a charm. Alot of the skill is knowing linux, js, and some of these other languages that other older control engineers don't know. When I have to go back and setup Factory Talk SE/ME and that kinda stuff its a total drag, just so limiting and I would argue fragile. So many hours on the Rockwell Tech Connect looking for some dumb patch so SQL will work, stuff like that is infuriating. I'm hoping in the future that companies like OPTO etc will turn the industry upside down and as older CE's retire the younger gen will adopt more powerful languages that we can more easily interface with web tools. I mean its only gonna get crazier with machine learning and data analytics, ladder logic was designed for relay type logic. We are expected to do way crazier stuff and we are doing round peg in square hole type programming constantly.o Its really cumbersome. I get why it works, but I long to see it die. Its moving slowly but as long as I we can prove the concepts and save money I think we have a chance.
I don't want to divert this thread, because I have very limited experience with SCADA applications, but I do see some real parallels with military systems where I have worked extensively. Requirements for reliability, maintainability, "idiot-proofing," etc. in mission (life) critical settings have driven these systems down a peculiar path that leaves them dependent on designs and protocols rarely used in the commercial or consumer world. The hardware is often expensive and a generation or two behind the current technology. Attempts to bring "commercial off-the-shelf" hardware and software to the field have met with mixed success, mainly (I think) because it is hard to find engineers who understand both the military and civilian sides of the house. I hope that SCADA is an easier nut to crack.
@drmibell I would love to hear you elaborate on this further. The real irony in the SCADA stuff is the paranoia surrounding security has actually left the protocols and design of the systems extremely open to attack. Old un-secured protocols really only have security-through-obscurity. Many of these systems were never intended to be put on the internet and therefore were never designed to have security features that can withstand todays attack vectors. I'd imagine that military systems suffer the same fate, albeit probably less so in the security department but I could see many of the same pitfalls. SCADA systems run the countries infrastructure and therefore could be considered a military threat so I think its a valid discussion. Water, waste water, power grid, gas, oil, transportation, etc are all valid attack targets and many run on garbage systems. It is hard to find people who understand both sides because the information is not shared outside of the communities. I personally don't think our current strategy of security-through-obscurity is sound and will eventually backfire. SCADA is probably easier to crack than the military stuff though I'd imagine. We can't launch missles, only turn off the water and power
I would say it really already has failed -- so many PLC vulnerabilities on ICS-CERT. And so many people leaving default passwords on things or exposing unsecured things to the internet -- a nice edge computer with a modern LTS Linux kernel, a firewall, and OpenVPN with appropriate ciphers blows most SCADA stuff out of the water security-wise.
Totally hear you on the round peg in square hole type programming. Ladder logic doesn't lend itself well to modern requirements of the smartphone-enabled world. Very cool that you're using the BBB in that type of environment -- just goes to show that embedded computers can go where PLCs can.
This is probably not the best place to go into details, but if you like I can send a few links for you to follow. The program I know best was called Acoustic Rapid COTS Insertion. It converted US Navy submarines to ruggedized versions of commercial computing hardware running almost-standard operating systems. This gave a huge increase in computing power (compared to military-standard computers) so that state-of-the-art signal-processing algorithms could be run. The program was hailed as a great success, but it ran afoul of almost every possible bit of the bureaucracy, from acquisition, to operational testing, to encryption certification. The approach got some traction in a few other programs where the payoff was worth the pain, but it never came close to replacing the old way of doing business.
I’m not alone!!
I want to see node-red fully take the place of the normal HMI. Alarm lists & history is all we are missing.
It could then be packaged as node-red-hmi with tailored nodes for hmi development.
I wouldn’t touch another Panelview.
@drmibell Yea, sure send me some links. This is interesting. Old ways die hard
netwingduck@gmail.com
@knolleary Well done to you and your team mate.