Very good presentation. I think the multitasking issues you experienced with the RPi are likely just programming issues since update threads need to work with atomic operations or have appropriate mutexes to guard against the issues mentioned. An important consideration that wasn't mentioned is maintainability. Design an industrial system with a PLC and any competent engineer should be able to maintain, diagnose and repair the system for the next 20+ years. With many controllers, the PLC will still be available, or at least a plugin-compatible one. If you need to replace the raspberry pi 4 board in 20 years, finding a replacement is likely to be a major issue.
Nice video. The raspberry PI has neither UL nor CE marking for industrial environments. It is also susceptible to transients and electrical noise. It is entirely unsuitable for industrial environments. This video notes that there are interfacing issues (no 24vdc for example) which is a practical problem because signaling is a science in itself and industrial sensors are generally 24vdc. A very big issue is that python or any other language supported on the Raspberry PI are poor choices for the factory floor. There are reasons that object oriented programming and dynamic memory allocation are very very little used on factory floors. These introduce risks that can result in instability, increasing the likelihood of malfunction, equipment or process damage or injury. The typical python/c++ programmer has zero experience with factory operations. The typical PLC programmer has zero experience with python/c++. Switching roles is possible but it takes a while to learn the rules. Another reason to avoid the Raspberry PI is that availability of equipment is variable and generally shorter term. Allen Bradley and Automation Direct and Siemens continue to sell parts for old systems for twenty years or more. Raspberry PI was intended to be an expendable learning device that I would have loved to have as a child. There are some expansion modules that allow it to be used in pseudo industrial environments, such as the Measurement Computing "pi hats". The best thing a PI has to offer is the OS (Linux and Win10 IOT), which can be used for things a PLC couldn't dream of. PLCs have been designed from the beginning for industrial control. They now commonly integrate safety functions, advanced robotic control, very deterministic operation and this; real actual technical support with a phone number and people who know what they are doing in industrial environments that speak your language fluently. If you are building an industrial project, it is foolish to use a Raspberry PI as the primary controller because it is easily questioned by any inspector or discerning customer. If you are building a proof of concept or experiment, you might be wasting your time because PLCs are so much easier to implement. If you are a student using a PI or Arduino to make a beer machine for your senior class project, remember that real beer makers use PLCs and that your skills developed using the PI/Arduino are only somewhat transferable to industry. The languages to know are IEC61131 (ladder logic, structured text, and to a lesser extent function block diagram, and the awful instruction list language if you live in Germany). BTW, good luck finding a college that teaches hard core programming in any of these languages. Luckily they are easy to learn like python, and even easier to troubleshoot because nearly every PLC has "online editing" which is the ability to perform debug and actual changes to the code while the system is running. Some PLCs even allow multiple programmers to make edits to the code simultaneously. It is interesting to note that the least expensive PLC on the market is the one in this video, the Automation Direct Click PLC.
@Howard Bleier The kunbus product looks awesome! It seems that it's purpose is to bring PC-based computing to the industrial landscape. It's moderate pricing could make it viable for some projects. I bet most people wouldn't be using it with Codesys or OpenPLC. If I were to use it, I might use it with OpenPLC and OpenCV to get the lowest cost controller and vision appliance. But that would probably be brutal to implement, lol. It will be interesting to see what they will offer this year at Nuremburg. Thanks for sharing!
Great exhaustive comments on your first message and very good points raised for whoever is thinking about Pi as a PLC. It does have very big differences in what is should be used for, even though it might not be very clear for someone unfamiliar with either disciplines (industrial automation vs programming). As you say, it really shouldn't be considered as the *primary* controller for an industrial machine, but could it have a place somewhere i wonder? Perhaps as a HMI / SCADA that might be more flexible to program? (I'm researching this now for a project)
I did in the past a semi-industrial project with an arduino. What I found is that the hydraulic coils did create instabilities, due to magnetism and transients. I did the same system back with a Click and it works very well!
Very insightful video. Another crucial differentiator between the two computers (because that's what they are) is long-term runtime reliability especially operating in hazardous environments. With that said though, I can see a future where Pi-based automation systems are introduced because of just how powerful and flexible these SBCs are. But the engineering and certification processes would probably end up costing as much as developing a dedicated, segragated PLC.
After taking industrial pneumatics and hydraulics courses and working with python my thoughts would be that the pi would work great with a PLC not a replacement. A PLC and ladder logic are so focused and the pi just lacks specific industrial traits. Together you'd have a terminator. Thanks for the insight.
satchell78, We are glad you enjoyed the video. Yes, they both have their applications in the industrial environment. Many engineers and self taught builders find the PI and Arduino products easy and inexpensive but lack the industrial ruggedness which is why we created the Productivity Open products.
This is a very useful video. It would be great if you'd upload the code & PLC project for both to Github or just your website so we can compare. I imagine the ladder logic is much simpler and more straightforward. Also, it would be useful to see the BOM for each because while the Pi is very inexpensive, costs start to add up once you add all the parts in you custom board, cases, transformers, connectors, optical isolators, etc.
SP30305ATL, We have both projects and will upload them to our video site. I will let you know once they are live. Thank you for asking and for your interest in both products.
I'm not sure this was a fair comparison. You're staying that interrupt requests from the raspberry pi make the peripherals slower, but you didn't mention if the PLC was using an HMI. Microsoft is notorious for not allocating interrupt requests to the peripherals.
What about using codesys on a raspberry pi? I'll have to double-check the feature-set, but the ability to have instruction list, ladder, structure text, and CFC all IEC I think makes the raspberry pi superior Also I think you can use profinet devices with the raspberry pi
Jonathan, we have seen many options for industrial interfacing devices to the Raspberry Pi (ProfiNet, Ethernet/IP etc.). As stated in another comment, the video was created by a third party with a non-biased opinion to show how each product could potentially provide a solution depending on application and programming experience. While some programmers may find it easier to program single board controllers, others, like those in the industrial field, may find PLCs are better suited for the application. The open source concept used with single board products allows for many control options and we at AutomationDirect are intrigued with the possibilities these controllers offer. For obvious reasons, we lean more towards PLCs but we do recognize the potential of single board controllers. If you are not familiar with our low cost CLICK PLC, we encourage you to review them and their free software and don’t forget to subscribe to our channel so you will be notified of any new products/videos we release 😉.
My 2 cents.. I am a beginner.. There is one main question when choosing Do you want a project or do you want to be done? My home automation runs of a Ethernet Click and node-red on the Pi. If you don't want to deal with project boards and have your actual project, home automation in my case the decision is easy. I do not live in the the US and I got my click fairly expensive but.. automation and industrial ready..
Thanks for the comment Dawid. Not being suited for industrial environment and time consumed is the trade off with some of the Open source technologies. That is why the New Productivity Open Industrial Arduino compatible controller is so impressive. Please see more info on this platform here: www.automationdirect.com/adc/shopping/catalog/programmable_controllers/open_source_controllers_(arduino-compatible)
instead just using python there are more PLC like softwares for the pi. Speedplc and codesys, I would love to see a video done with these softwares instead for a more comparable review!
Great video! I personally think you’ll see more raspberry pi’s in factory automation. Imagine you stock pile raspberry single board computers instead of very expensive PLC units. You can keep production cost down and having replacement parts on hand wouldn’t cost thousands of dollars or special tooling and an engineer with a specific background in name brand automation software. I also love the community behind raspberry pi! Keep up the great work.
I am not sure about using raspberry PI for industrial applications. We tried to do a automation project using a beaglebone computer. Firstly We had to add an extra layer of protective electronics for the IOs, which increased the cost of the product. Secondly we had provide backup power so the Beaglebone operating system (Debian) would shutdown correctly. Unfortunately this too added to the cost of the product. The final product worked as per specs, but the whole lot of extra hardware made the system unwieldy and precarious. I would prefer a PLC anytime over a raspberry, even a cheap Chinese made PLC works very well. It's possible to use raspberry style computers but not in its present form
Raspberry PI is failed UK government project but Uncle Sam is trying to shove it down our kids brains. It's useless as a SOC computer(too expensive), controller (too slow to respond to real time events). Python is a brain dead outdated language that will put American kids behind for generations being pushed by Google's limitless money. My 5th grader is learning C instead to control the robots.
This is a fair comparison of the two options. Well done. When I want to control my Christmas tree lights: RASBERRYY PI When I want to control something important: PLC Easy :-)
Nothing wrong with this video or the statements made. However, Raspberry Pies are probably more used for home automation or hobby projects than industrial automation tasks, although I've seen RasPi based PLCs as well. Still, for many makers, most of the PLCs out there are just prohibitively expensive. Even the programming software alone for a Siemens S300 breaks the bank for most home automation projects. Most makers have ample time at their hands, but no equivalent amount of money. I recently took note of the Click (Plus) PLC range, and fell in love with it, as the pricing makes it accessible for my home automation. So I trade money for time and have less hassle. But if a maker is out for saving the maximum amount of money and doesn't care about time, the RasPi or an Arduino is almost impossible to beat. Just a few features I'd love to have on a Click, which are e.g. modules for inexpensive Dallas 1-wire temperature sensors or DHT22-AM2302 temp and humidity sensors. Purchasing a $180 RTD input module, and another $60 for 4 sensors is much less value for the money than getting an Arduino Uno clone, Ethernet shield and 4 DS18B20 for a total of $30.
An Arduino starter kit from Spakfun is about $100. A Do More starter kit from AD is about $500. Not apples to apples, but one definetely an easier entry point.
I've always been disappointed with Raspberry Pi IO. It always just sems so limiting, you'd think common industrial sensor signals would be a priority on this thing. We use a couple in a hotel to view a couple IP cameras over the normal cable television. They work great for that (when they're working that it lol)
pi4 with codesys runtime and linux rt patch with a constant very small jitter clocked a 2.0Ghz tested, usb3.0 ssd boot beats any plc you throw at in any application as long as you provide suitable industrial housing respecting IP standards.
No it doesn't, a raspberry pi cannot handle several hundred or thousand i/o signals, it cannot handle realtime motion control. An industrial enclosure will not turn a raspberry pi into a 12 axis servo controller however pretty, big and shiny the enclosure may be. USB3 doesn’t like noise on its data lines - hence specific cables for the higher data rate. Useless in industrial situations where electrical noise is prevalent. I have seen desktop computers freeze and lock up due to a usb mouse cable that was in close proximity to a motor VFD. When the motor started anything usb connected would drop out, become undetectable and cause the the computer to lock up until the drive power was removed and the usb items were re inserted.
Nate, We will make note of your request for a video on Data Acquisition for the Raspberry Pi and a CLICK PLC. If you create one, please let us know, we love sharing industrial resolutions with our AutomationDirect products.
I agree that for industrial automation the right tool is a PLC (input/output signal conditioning built in, screw terminals, resistance ho harsh industrial conditions, simple programming language every "dumb" electrician can understand). On the other hand, for a student an Arduino/ESP32/Raspberry PI is the way to go. Much lower price, endless possibilities to interface to interesting advanced sensors (available for peanuts from China), and above all you use a REAL programming language (typically C/C++) that is the de facto standard in the IT industry. If you know C++ you can do everything (not necessarily in industrial automation), and also learn to program PLCs in a weekend by your own. On the other hand the electrician who only knows ladder, or other PLC programming "languages", will stay at that level for the rest of his life (fewer career opportunities).
Thanks for the video and your efforts, but why python ? Bare metal C, and just a little tweaked kernel (such as real time patch and removing unnecessary kernel code) will not only make Pi much better but also its parameters can be changed while it is working!! (a linux process can be notified while it is working that a conf file has been changed with many different ways such as inotify, sockets, etc..), it can also report everything in real time, it can be updated instantly. But for reliability and long time stable working hardware must be well protected and that can be challenging.
Pi can replace expensive processors. You could use codesys on raspberry Pi . It support motion and most of the industrial protocols you can imagine. Codesys for those who have worked in plc programming is one of the best programming environment. Used by abb, wago, beckoff plc's
May I add, that one could just for a thought experment use a more "plc like" operating system for a singe board computer and use proper "real time capable" programming techniques and maybe additional hardware to conform to indutry standard? At it's inner core a plc is going to look very close to a single board comupter like a raspberry pi. The inputs and outputs of - i.e. a siemens cpu - are unlikely to conform to indutry standards when viewd in isolation of the circuitry inside a plcs housing. A traditional plc alredy contains the circuitry needed for the nessasary conversions. I am only capable of speculation with the following things, but as cycle-times are a thing on many plcs, i doubt my hypothetical assumptions. Lets say, a typical plc processor is at its core an fpga or cpld device (programmable hardware) that is combined with a classical processor running an operating system for monitoring and asyncrounus communication tasks. The simple logic could be compiled to run on the fpga part in true hardware fashion and infinite speed. Assuming that you could run pretty much als much logic on a plc as you please. The only thing that would require any action of a processor would be monitoring of the plc and maybe networking. In that case only the smallest and most basic plcs would worry about anything timing related at all. So I suggest that a a plc is - very simplified - nothing more than a raspberry pi with a very specialized operating system and some levelshifters. So if someone really wanted, one could contruct the kind of circuit that a plc includes around a raspberry pi and would get exactly the same as a plc. This comment was only meant as a note, so nobody would think that a raspberry pi is inherently inferior or or plcs perform magic to do things the device used to view thic comment could not in theory. Neither was my internet to talk badly or tarnish a really good video. 😉
Python is a great and simple language, but i personally would have chosen differently. I'd likely chosen c++ or rust and would have tried to implement interrupt code.
There is some really good points here. Software and programming could be implemented on Raspberry Pi given the right platform. Hardware is where this difference has been really made between the industrial nature of a Raspberry Pi and a PLC, example Circuit isolation and other considerations. These differences has and will become less drastic as more industrial nature Raspberry Pi and Arduino become more common. For example the Productivity Open that mixes Arduino and industrial I/O.
@@automationdirect I liked your video, but really wanted to add these things with more details, so nobody get a wrong idea. I built a fairly small framework using include os, that serves the states alongside a primitive web interface for start, stop, shutdown and reboot and uses buffers in memory to ensure everything is updated at the same time. Admittedly c++ might not be the most convinient thing, but it could serve as a base for expansion if I need it. The repository isn't public for now, because I had not enough time to polish the experience using it....
use raspberry, arduino etc as inputs for real PLCs, thats how i solved the home automation issue. siemens lacks of http and other stuff and vice versa. Raspberry can fetch data from internet and show "green light" to siemens Logo! to execute more code.
Great point Tomi. Adding a microcontroller (like Arduino or Raspberry) to your PLC control system as a auxiliary embedded device, is a good way to merge the benefits of both platforms.
You forgot the cost of a click plc vs raspberry pi. It's like me trying to convince you to buy a Ferrari over a toyota. Well theres about half a million dollars difference but the Ferrari is better.
The monies involved with Click is small scale compared to the world of Rockwell Automation. I wish great success with your projects at hand. Peace brother. ✌🏽
Oh my codesys on the raspberry pi for far more advanced plc programming than anything automation direct currently offers,......then use wago or bechkoff expansion and communication modules or if you're up to it any modbus, canbus,profibus, profinet, eathecat, industrial ethernet device you wish ....just use the raspberry pi as the controller n u get a suppeprior plc programming environment.....but its not certified for industrial use but its awwwwwssssoooooommmmmeeee
JD JD, Thank you. Make sure to check out more of our videos at www.automationdirect.com/videos/home We have over 1400 videos on the automation products we offer here at Automationdirect.com
Really bad definition of Python, procedural is only one of the different programming paradigms available in Python, you can also use a realtime patch for RPi3 among other options to overcome realtime issues, this comparison is poor and it seems that the only effective comparison was the capacity of the creator to program in ladder logics against his own knowledge of Python.
PLC is the only way to go.... depending on reliability for million dollar machinery or some low budget science fair? The customer will always blame OEM and or the Programer for the failure. I never try to cut ✂️ the dollar before letting my customer's know what they are talking about... sacrifice reliability in safety standards is a big deal. Windows 10 on PC is still locking up, lol. Maybe that's why 11 is now out... Good topic... its just that important. Thank's
Adam, thanks for the comment. Many associate Raspberry Pi and Arduino both as being single board controllers and both potential candidates for controls in industrial environments. There is a video comparing an Arduino and a CLICK PLC, hosted on ControlDesigns channel: ua-cam.com/video/9KJ-bQ2TPlQ/v-deo.html .
Very good presentation. I think the multitasking issues you experienced with the RPi are likely just programming issues since update threads need to work with atomic operations or have appropriate mutexes to guard against the issues mentioned. An important consideration that wasn't mentioned is maintainability. Design an industrial system with a PLC and any competent engineer should be able to maintain, diagnose and repair the system for the next 20+ years. With many controllers, the PLC will still be available, or at least a plugin-compatible one. If you need to replace the raspberry pi 4 board in 20 years, finding a replacement is likely to be a major issue.
Thank you for the comment Michael. Very good points.
Awesome video! Someone took our "you need a PLC" rant and turned it into a retroencabulator style video. Much appreciated, will absolutely be sharing.
Thanks Dennis!
Nice video.
The raspberry PI has neither UL nor CE marking for industrial environments. It is also susceptible to transients and electrical noise. It is entirely unsuitable for industrial environments. This video notes that there are interfacing issues (no 24vdc for example) which is a practical problem because signaling is a science in itself and industrial sensors are generally 24vdc. A very big issue is that python or any other language supported on the Raspberry PI are poor choices for the factory floor. There are reasons that object oriented programming and dynamic memory allocation are very very little used on factory floors. These introduce risks that can result in instability, increasing the likelihood of malfunction, equipment or process damage or injury.
The typical python/c++ programmer has zero experience with factory operations. The typical PLC programmer has zero experience with python/c++. Switching roles is possible but it takes a while to learn the rules.
Another reason to avoid the Raspberry PI is that availability of equipment is variable and generally shorter term. Allen Bradley and Automation Direct and Siemens continue to sell parts for old systems for twenty years or more.
Raspberry PI was intended to be an expendable learning device that I would have loved to have as a child. There are some expansion modules that allow it to be used in pseudo industrial environments, such as the Measurement Computing "pi hats". The best thing a PI has to offer is the OS (Linux and Win10 IOT), which can be used for things a PLC couldn't dream of.
PLCs have been designed from the beginning for industrial control. They now commonly integrate safety functions, advanced robotic control, very deterministic operation and this; real actual technical support with a phone number and people who know what they are doing in industrial environments that speak your language fluently.
If you are building an industrial project, it is foolish to use a Raspberry PI as the primary controller because it is easily questioned by any inspector or discerning customer. If you are building a proof of concept or experiment, you might be wasting your time because PLCs are so much easier to implement. If you are a student using a PI or Arduino to make a beer machine for your senior class project, remember that real beer makers use PLCs and that your skills developed using the PI/Arduino are only somewhat transferable to industry. The languages to know are IEC61131 (ladder logic, structured text, and to a lesser extent function block diagram, and the awful instruction list language if you live in Germany). BTW, good luck finding a college that teaches hard core programming in any of these languages. Luckily they are easy to learn like python, and even easier to troubleshoot because nearly every PLC has "online editing" which is the ability to perform debug and actual changes to the code while the system is running. Some PLCs even allow multiple programmers to make edits to the code simultaneously. It is interesting to note that the least expensive PLC on the market is the one in this video, the Automation Direct Click PLC.
Dennis, great comment and description of industrial usage. :)
@Howard Bleier The kunbus product looks awesome! It seems that it's purpose is to bring PC-based computing to the industrial landscape. It's moderate pricing could make it viable for some projects. I bet most people wouldn't be using it with Codesys or OpenPLC. If I were to use it, I might use it with OpenPLC and OpenCV to get the lowest cost controller and vision appliance. But that would probably be brutal to implement, lol. It will be interesting to see what they will offer this year at Nuremburg. Thanks for sharing!
Great exhaustive comments on your first message and very good points raised for whoever is thinking about Pi as a PLC. It does have very big differences in what is should be used for, even though it might not be very clear for someone unfamiliar with either disciplines (industrial automation vs programming).
As you say, it really shouldn't be considered as the *primary* controller for an industrial machine, but could it have a place somewhere i wonder? Perhaps as a HMI / SCADA that might be more flexible to program? (I'm researching this now for a project)
Lol what a comment! 👍🤣
I did in the past a semi-industrial project with an arduino. What I found is that the hydraulic coils did create instabilities, due to magnetism and transients. I did the same system back with a Click and it works very well!
Yes, Those are some of the subtle differences folks need to watch out for.
Very insightful video. Another crucial differentiator between the two computers (because that's what they are) is long-term runtime reliability especially operating in hazardous environments. With that said though, I can see a future where Pi-based automation systems are introduced because of just how powerful and flexible these SBCs are. But the engineering and certification processes would probably end up costing as much as developing a dedicated, segragated PLC.
After taking industrial pneumatics and hydraulics courses and working with python my thoughts would be that the pi would work great with a PLC not a replacement. A PLC and ladder logic are so focused and the pi just lacks specific industrial traits. Together you'd have a terminator. Thanks for the insight.
satchell78, We are glad you enjoyed the video. Yes, they both have their applications in the industrial environment. Many engineers and self taught builders find the PI and Arduino products easy and inexpensive but lack the industrial ruggedness which is why we created the Productivity Open products.
This is a very useful video. It would be great if you'd upload the code & PLC project for both to Github or just your website so we can compare. I imagine the ladder logic is much simpler and more straightforward. Also, it would be useful to see the BOM for each because while the Pi is very inexpensive, costs start to add up once you add all the parts in you custom board, cases, transformers, connectors, optical isolators, etc.
SP30305ATL, We have both projects and will upload them to our video site. I will let you know once they are live. Thank you for asking and for your interest in both products.
I'm not sure this was a fair comparison. You're staying that interrupt requests from the raspberry pi make the peripherals slower, but you didn't mention if the PLC was using an HMI. Microsoft is notorious for not allocating interrupt requests to the peripherals.
Thanks for the comment. Good points!
What about using codesys on a raspberry pi?
I'll have to double-check the feature-set, but the ability to have instruction list, ladder, structure text, and CFC all IEC I think makes the raspberry pi superior
Also I think you can use profinet devices with the raspberry pi
Jonathan, we have seen many options for industrial interfacing devices to the Raspberry Pi (ProfiNet, Ethernet/IP etc.). As stated in another comment, the video was created by a third party with a non-biased opinion to show how each product could potentially provide a solution depending on application and programming experience. While some programmers may find it easier to program single board controllers, others, like those in the industrial field, may find PLCs are better suited for the application. The open source concept used with single board products allows for many control options and we at AutomationDirect are intrigued with the possibilities these controllers offer. For obvious reasons, we lean more towards PLCs but we do recognize the potential of single board controllers. If you are not familiar with our low cost CLICK PLC, we encourage you to review them and their free software and don’t forget to subscribe to our channel so you will be notified of any new products/videos we release 😉.
My 2 cents.. I am a beginner.. There is one main question when choosing Do you want a project or do you want to be done? My home automation runs of a Ethernet Click and node-red on the Pi. If you don't want to deal with project boards and have your actual project, home automation in my case the decision is easy. I do not live in the the US and I got my click fairly expensive but.. automation and industrial ready..
Thanks for the comment Dawid. Not being suited for industrial environment and time consumed is the trade off with some of the Open source technologies. That is why the New Productivity Open Industrial Arduino compatible controller is so impressive. Please see more info on this platform here: www.automationdirect.com/adc/shopping/catalog/programmable_controllers/open_source_controllers_(arduino-compatible)
instead just using python there are more PLC like softwares for the pi. Speedplc and codesys, I would love to see a video done with these softwares instead for a more comparable review!
Thanks for the idea Tom!
Great video! I personally think you’ll see more raspberry pi’s in factory automation. Imagine you stock pile raspberry single board computers instead of very expensive PLC units. You can keep production cost down and having replacement parts on hand wouldn’t cost thousands of dollars or special tooling and an engineer with a specific background in name brand automation software. I also love the community behind raspberry pi! Keep up the great work.
Thanks for the comment.
I am not sure about using raspberry PI for industrial applications. We tried to do a automation project using a beaglebone computer.
Firstly We had to add an extra layer of protective electronics for the IOs, which increased the cost of the product. Secondly we had provide backup power so the Beaglebone operating system (Debian) would shutdown correctly. Unfortunately this too added to the cost of the product. The final product worked as per specs, but the whole lot of extra hardware made the system unwieldy and precarious.
I would prefer a PLC anytime over a raspberry, even a cheap Chinese made PLC works very well. It's possible to use raspberry style computers but not in its present form
Raspberry PI is failed UK government project but Uncle Sam is trying to shove it down our kids brains. It's useless as a SOC computer(too expensive), controller (too slow to respond to real time events). Python is a brain dead outdated language that will put American kids behind for generations being pushed by Google's limitless money. My 5th grader is learning C instead to control the robots.
This is a fair comparison of the two options. Well done.
When I want to control my Christmas tree lights: RASBERRYY PI
When I want to control something important: PLC
Easy :-)
Thanks Factory! If you want to learn more about our other products go to www.automationdirect.com/videos we have over 1000 videos to watch!
Nothing wrong with this video or the statements made. However, Raspberry Pies are probably more used for home automation or hobby projects than industrial automation tasks, although I've seen RasPi based PLCs as well.
Still, for many makers, most of the PLCs out there are just prohibitively expensive. Even the programming software alone for a Siemens S300 breaks the bank for most home automation projects. Most makers have ample time at their hands, but no equivalent amount of money.
I recently took note of the Click (Plus) PLC range, and fell in love with it, as the pricing makes it accessible for my home automation. So I trade money for time and have less hassle. But if a maker is out for saving the maximum amount of money and doesn't care about time, the RasPi or an Arduino is almost impossible to beat.
Just a few features I'd love to have on a Click, which are e.g. modules for inexpensive Dallas 1-wire temperature sensors or DHT22-AM2302 temp and humidity sensors. Purchasing a $180 RTD input module, and another $60 for 4 sensors is much less value for the money than getting an Arduino Uno clone, Ethernet shield and 4 DS18B20 for a total of $30.
Thanks for your comments. I will pass the information on to the developers.
An Arduino starter kit from Spakfun is about $100. A Do More starter kit from AD is about $500. Not apples to apples, but one definetely an easier entry point.
Thank you for the comment. This video compares a Raspberry Pi to a CLICK PLC. AutomationDirect offers CLICK PLC's starting at $69.00.
I've always been disappointed with Raspberry Pi IO. It always just sems so limiting, you'd think common industrial sensor signals would be a priority on this thing. We use a couple in a hotel to view a couple IP cameras over the normal cable television. They work great for that (when they're working that it lol)
pi4 with codesys runtime and linux rt patch with a constant very small jitter clocked a 2.0Ghz tested, usb3.0 ssd boot beats any plc you throw at in any application as long as you provide suitable industrial housing respecting IP standards.
Thanks for the comment, Make sure to check more of our videos www.automationdirect.com/videos/home
No it doesn't, a raspberry pi cannot handle several hundred or thousand i/o signals, it cannot handle realtime motion control. An industrial enclosure will not turn a raspberry pi into a 12 axis servo controller however pretty, big and shiny the enclosure may be.
USB3 doesn’t like noise on its data lines - hence specific cables for the higher data rate. Useless in industrial situations where electrical noise is prevalent. I have seen desktop computers freeze and lock up due to a usb mouse cable that was in close proximity to a motor VFD. When the motor started anything usb connected would drop out, become undetectable and cause the the computer to lock up until the drive power was removed and the usb items were re inserted.
Now make a video showing how to use the raspberry pi as a DAQ for the Click.
Nate, We will make note of your request for a video on Data Acquisition for the Raspberry Pi and a CLICK PLC. If you create one, please let us know, we love sharing industrial resolutions with our AutomationDirect products.
I agree that for industrial automation the right tool is a PLC (input/output signal conditioning built in, screw terminals, resistance ho harsh industrial conditions, simple programming language every "dumb" electrician can understand).
On the other hand, for a student an Arduino/ESP32/Raspberry PI is the way to go. Much lower price, endless possibilities to interface to interesting advanced sensors (available for peanuts from China), and above all you use a REAL programming language (typically C/C++) that is the de facto standard in the IT industry.
If you know C++ you can do everything (not necessarily in industrial automation), and also learn to program PLCs in a weekend by your own. On the other hand the electrician who only knows ladder, or other PLC programming "languages", will stay at that level for the rest of his life (fewer career opportunities).
A lot of great points! Thanks for the comment.
Thanks for the video and your efforts, but why python ? Bare metal C, and just a little tweaked kernel (such as real time patch and removing unnecessary kernel code) will not only make Pi much better but also its parameters can be changed while it is working!! (a linux process can be notified while it is working that a conf file has been changed with many different ways such as inotify, sockets, etc..), it can also report everything in real time, it can be updated instantly. But for reliability and long time stable working hardware must be well protected and that can be challenging.
Thanks for the comment Onur. Great food for thought for anyone making this comparison
Pi can replace expensive processors. You could use codesys on raspberry Pi . It support motion and most of the industrial protocols you can imagine. Codesys for those who have worked in plc programming is one of the best programming environment. Used by abb, wago, beckoff plc's
Codesys controllers are expensive,for simple project ,why not use cheap PLC?
Thanks for the feedback Joel!
May I add, that one could just for a thought experment use a more "plc like" operating system for a singe board computer and use proper "real time capable" programming techniques and maybe additional hardware to conform to indutry standard?
At it's inner core a plc is going to look very close to a single board comupter like a raspberry pi. The inputs and outputs of - i.e. a siemens cpu - are unlikely to conform to indutry standards when viewd in isolation of the circuitry inside a plcs housing.
A traditional plc alredy contains the circuitry needed for the nessasary conversions.
I am only capable of speculation with the following things, but as cycle-times are a thing on many plcs, i doubt my hypothetical assumptions. Lets say, a typical plc processor is at its core an fpga or cpld device (programmable hardware) that is combined with a classical processor running an operating system for monitoring and asyncrounus communication tasks.
The simple logic could be compiled to run on the fpga part in true hardware fashion and infinite speed.
Assuming that you could run pretty much als much logic on a plc as you please. The only thing that would require any action of a processor would be monitoring of the plc and maybe networking.
In that case only the smallest and most basic plcs would worry about anything timing related at all.
So I suggest that a a plc is - very simplified - nothing more than a raspberry pi with a very specialized operating system and some levelshifters. So if someone really wanted, one could contruct the kind of circuit that a plc includes around a raspberry pi and would get exactly the same as a plc.
This comment was only meant as a note, so nobody would think that a raspberry pi is inherently inferior or or plcs perform magic to do things the device used to view thic comment could not in theory.
Neither was my internet to talk badly or tarnish a really good video. 😉
Python is a great and simple language, but i personally would have chosen differently. I'd likely chosen c++ or rust and would have tried to implement interrupt code.
There is some really good points here. Software and programming could be implemented on Raspberry Pi given the right platform. Hardware is where this difference has been really made between the industrial nature of a Raspberry Pi and a PLC, example Circuit isolation and other considerations. These differences has and will become less drastic as more industrial nature Raspberry Pi and Arduino become more common. For example the Productivity Open that mixes Arduino and industrial I/O.
@@automationdirect I liked your video, but really wanted to add these things with more details, so nobody get a wrong idea.
I built a fairly small framework using include os, that serves the states alongside a primitive web interface for start, stop, shutdown and reboot and uses buffers in memory to ensure everything is updated at the same time.
Admittedly c++ might not be the most convinient thing, but it could serve as a base for expansion if I need it.
The repository isn't public for now, because I had not enough time to polish the experience using it....
use raspberry, arduino etc as inputs for real PLCs, thats how i solved the home automation issue. siemens lacks of http and other stuff and vice versa. Raspberry can fetch data from internet and show "green light" to siemens Logo! to execute more code.
Great point Tomi. Adding a microcontroller (like Arduino or Raspberry) to your PLC control system as a auxiliary embedded device, is a good way to merge the benefits of both platforms.
You forgot the cost of a click plc vs raspberry pi. It's like me trying to convince you to buy a Ferrari over a toyota. Well theres about half a million dollars difference but the Ferrari is better.
The monies involved with Click is small scale compared to the world of Rockwell Automation. I wish great success with your projects at hand. Peace brother. ✌🏽
Thanks for the feedback. Make sure to check more of our videos and products on our website www.automationdirect.com/adc/home/home
Oh my codesys on the raspberry pi for far more advanced plc programming than anything automation direct currently offers,......then use wago or bechkoff expansion and communication modules or if you're up to it any modbus, canbus,profibus, profinet, eathecat, industrial ethernet device you wish ....just use the raspberry pi as the controller n u get a suppeprior plc programming environment.....but its not certified for industrial use but its awwwwwssssoooooommmmmeeee
Hi Christopher, yes, it all depends on your background and work environment.
That’s a cool video
JD JD, Thank you. Make sure to check out more of our videos at www.automationdirect.com/videos/home We have over 1400 videos on the automation products we offer here at Automationdirect.com
Complete noob question; Can arduinos come into play here?
They can come into the conversation but in a sperate way. Check out this video: ua-cam.com/video/KOjPn9LVcjI/v-deo.html
A very fair and good comparison
Appreciate it Rob!
Really bad definition of Python, procedural is only one of the different programming paradigms available in Python, you can also use a realtime patch for RPi3 among other options to overcome realtime issues, this comparison is poor and it seems that the only effective comparison was the capacity of the creator to program in ladder logics against his own knowledge of Python.
Thanks for your feedback Andres.
PLC is the only way to go.... depending on reliability for million dollar machinery or some low budget science fair?
The customer will always blame OEM and or the Programer for the failure. I never try to cut ✂️ the dollar before letting my customer's know what they are talking about... sacrifice reliability in safety standards is a big deal. Windows 10 on PC is still locking up, lol. Maybe that's why 11 is now out...
Good topic... its just that important.
Thank's
@AEI Automation Group, thanks for the comment. The reliability and track record of PLCs is hard to argue.
Click plc has no ST, just ladder....
That is correct. There is not structured text in the CLICK PLC.
You compared a computer to a PLC? Thats pretty stupid.
A better comparison would have been an arduino and a PLC.
Adam, thanks for the comment. Many associate Raspberry Pi and Arduino both as being single board controllers and both potential candidates for controls in industrial environments. There is a video comparing an Arduino and a CLICK PLC, hosted on ControlDesigns channel: ua-cam.com/video/9KJ-bQ2TPlQ/v-deo.html .