How can i use safely arduino at industry

[tlhsglm]

Hi fellows,

I want to use arduino in textile industry where there is lot of dust, micro textile flys, noise and electrostatic charges. This industry is for know, my purpess is to have is it suitable for industry?

I know ladder language to use PLC but they are so expensive and my knowledge in c, c++ is more deep. I want to make arduino roboust like plc.

I will use sensors, 220v elektric motors, also even cameras.

Dusts can be flatten away with nice casing but esd  and noise is confusing me. Is arduino can jump-miss lines on a complex long code? I know plc and arduino uses same chips but there is really no trusting arduino in industry why is that and how can we eliminate this?

[coromonadalix]

An arduino once programmed will act like any controller you want to use,  BUT   if you use it in a tough environement  like  surges spikes etc ....  it may act crazy or glitch etc ... like any interface

You have on google  some free arduino  plc projects ... the boards are better protected against some problems, and they provide some insulations  on the inputs and / or the outputs

Insulation interface for the motor controller   etc ...    using  solid state relays, opto coupled interfaces ...



The remaining problems would be you coding skills,  if you make some errors, the reliability of your project could be bad

[Siwastaja]

Arduino itself isn't the problem.

The problem is that you likely overestimate your skills, and underestimate the work (and experience!) needed to create a robust, reliable, industrial product.

It just happens that Leonardo Da Vinci did not use crayons, a carpenter won't use one of those plastic toy drills but a real drill, and a professional designer won't use Arduino. Not because they hate Arduinos, but because they do not need them, and find them insufficient and limiting.

Once you start gaining understanding and experience on building such products, it becomes obvious to you that you won't be using an Arduino.

[NiHaoMike]

Look up "Arduino PLC" and you'll find quite a few designs that solve the electrical robustness problem.

Then consider using PlatformIO instead of the "official" Arduino IDE which is more designed for beginners.

[wizard69]

I think you have this wrong, PLC's are the cheapest way to sustainable automation.   This especially if downtime gets charged by the minute.  The reason is the debugging tools and rapid hardware replacement.   Arduino's become black boxes that can only be serviced by the original programmer and sometimes not even then.   There is a huge contrast between a good, well supportted PLC system and an Arduino (which has no system).

[KE5FX]

Arduino itself isn't the problem.  The problem is that you likely overestimate your skills, and underestimate the work (and experience!) needed to create a robust, reliable, industrial product.

The people who "overestimate their skills" are the ones who don't ask for advice first.

Once you start gaining understanding and experience on building such products, it becomes obvious to you that you won't be using an Arduino.

Take a stroll through an EMS facility sometime.  You're in for a shock.  Those things are used in a lot of commercial products, along with other boards you wouldn't expect to see like Raspberry Pis.

[Apothus]

Take a stroll through an EMS facility sometime.  You're in for a shock.  Those things are used in a lot of commercial products, along with other boards you wouldn't expect to see like Raspberry Pis.

According to a rep I spoke with recently over a 3rd of all raspberry pi hardware sold is utilised within a commercial product. If you need an embedded Linux system in a product its often one of the cheapest and easiest if you're not ordering in quantities of 1000+

With Arduino, be careful about the libraries you use as they are often LGPL which means you probably need to publish some parts of the code to be compliant. Arduino are working on this and I'd suggest reaching out the them directly as they are putting a lot of effort into connecting with industry at the moment. Their Portenta line is an example of this.

WRT reliability. I would start by researching standard input protections, current limiting resistors, tvs diodes, zenner clamps, RC filters and the like. Watchdog timers to ensure your code does not lock up completely and design your circuit so that the equipment fails into a safe state in the event your microcontroller has a fault.

[elekorsi]

What you are saying just doesn't make any sense... Words "industry" and "plc to expensive" should not be in the same topic. As @wizard69 said, PLCs are the cheapest option for automation...
You can go with the custom electronics when you are a machine builder, producing n x 10000 of the same machines per year, backed up with an electronics design TEAM and probably they will not use an arduino...

[NiHaoMike]

Arduino's become black boxes that can only be serviced by the original programmer and sometimes not even then.

As compared to some other less popular microcontroller platform? Arduino is by far the most popular embedded platform. Maybe Micropython would take over following the popularity of Python scripting on test lab PCs, but that hasn't happened yet. (I personally think Python is a bit "heavy" for a microcontroller, but performance is rarely the biggest concern for microcontrollers used in test labs.)

You can go with the custom electronics when you are a machine builder, producing n x 10000 of the same machines per year, backed up with an electronics design TEAM and probably they will not use an arduino...

For automated testing in R&D test labs, I have seen a lot of Arduinos in use and also have used quite a few myself. For the common use case of bridging between a USB port on a PC and many low speed digital I/Os, it's hard to beat an Arduino Mega. It's quite common to see an actual Arduino board being used, particularly when the functionality can be done using off the shelf shields. (Try getting some custom test fixture boards designed and built and you'll understand.)

I have also seen Beaglebone Black being used as logic analyzers in a test lab. Its low cost and small size makes it trivial to have one on each setup and the Ethernet connection solves the remote access problem. (They do have a few "proper" logic analyzers but they're only used for high speed signals beyond the capabilities of a Beaglebone Black.)

[rfclown]

Takes me back to the first and only time I used a PLC. Someone I knew had his custom guitar body routing machine quit. The guy who designed the control system for it told him that he couldn't fix it, and "nobody does it like that anymore". He told my friend to get an PLC, so he bought this Mitsubishi thing. He couldn't figure it out, so I got a call. Took me a while to figure that the original system used diode-relay logic. Who would ever think of such a thing... but it was almost a work of art (but a rats nest). If he hadn't already bought the Mitsubishi thing, I would have used a microcontroller (probably a PIC at the time). I just googled "Mitsubishi PLC" and the first hit showed this:
https://www.radwell.com/Shop?source=GoogleShopping&IgnoreRedirect=true&ItemSingleId=4900371&utm_source=google&utm_medium=cpc&adpos=&scid=scplp4900371&sc_intid=4900371&gclid=EAIaIQobChMIlJHAjeuP9gIVAuTICh20AQ3fEAQYASABEgLr-PD_BwE
which looks just like what I used. This is NOT expensive. It was very easy use. It programmed in a graphical language. It seamlessly interfaced to the proximity switches, motors, buttons, knobs, etc. I don't know if this line of PLCs are still around (I have not since done this type of work), but I would use them again. I didn't have to worry about making a uC board that was robust in an industrial environment. The PLC just worked. It directly interfaced with high voltage I/O. It save a lot of time, and if anyone else wanted to deal with it in the future, they wouldn't have to figure out my assembly code (which is what I would have used for a PIC at that time).

[Berni]

Cheep PLCs are not all that expensive really when you factor in the extra stuff needed for the Arduino

The PLC comes already in a protective case that simply clips onto a DIN rail, this makes it easy to install in a industrial control cabinet that likely already has multiple DIN rails inside to hold all the relays, circuit breakers, power supplies, sensor converters and other stuff required to make the machine run. All the IO is also 12 or 24V or relay or similar that fits well with industrial machines and is designed to be super rugged. The PLC also typically includes a convenient display to show some status and buttons for adjusting things.

You have to factor in your time too.
Yes PLCs are sometimes pretty cumbersome to program, but the machines they run typically require pretty simple control anyway. So all you need to do is buy one, wire it in and make a program for it. No extra electronics development needed and if a big lightning strike comes and blows it all up you can easily just pick up a new PLC from the store and stick it in. Even someone who has never seen the system can fix it by just replacing the PLC as long as you kept the software somewhere (I like leaving SD cards with firmware inside my stuff)

[nali]

Cheep PLCs are not all that expensive really when you factor in the extra stuff needed for the Arduino

This^.

Reminds me of a former company where I was the sole HWeng in a bunch of SW & management types, where one of the applications was fitting mobile PCs on to buses, which is a pretty horrible environment in all respects. I lost count of the number of times someone left a magazine clipping on my desk "NEW $10 computer runs linux" (actually a SoC) or "Can't we use a Raspberry PI?" etc.

Anyways after I left they decided to try the RPi route. After commissioning someone to add a custom enclosure, add an automotive grade buck converter, add interfaces yadda yadda they ended up with something i) More expensive ii) Less reliable iii) Slower and iv) Not type approved.