I have an idea for a small device to be used on a machine in a tier 2 Automotive Plant.
The device will utilize two small steppers, read an analog signal, and communicate with a PLC.
In general do people use the Arduino products for hobby projects and then move on to something else for professional products, or is the Arduino product suitable for a professional finished device?
Thanks in advance for your comments and time.
I don't know what a Tier 2 plant is and what that might mean for certification and safety of components.
One of the elements of the Arduino's success is yes you can do that. You can develop your code on inexpensive hardware without having to buy a programmer or software licence. Then you can design your own hardware using the same chip and the same code still runs on that "professional" hardware.
There are probably hundreds of thousands of industrial machines out there now, with a small Arduino like a Teensy or Mini permanently soldered onto the board.
My current project is to control a spectrometer using the Arduino Mega, among other things. It is for a professional application (mainly to compare my data to the data of better spectrometers). I'd say yes, it can be a professional finished piece if you want it to be.
I'd say the important thing for industrial applications is that it be rugged (electrically and mechanically), **reliable, and serviceable. **
Most solid state electronics is reliable as long as it's protected from electrical/environmental/mechanical stress. It's more about how you build the thing more than what processor & electronics you use. ...Although, they might tease you if you use a Arduino or PIC or other "hobby" microcontroller. 
There is an electrically beefed-up version of the Arduino called the [u]The Ruggeduino[/u].
Serviceability is always a concern with anything custom made... If the guy that built it goes-away, maybe nobody can fix it. Or, maybe the guy who built it has gone home for the day, etc.
It helps to keep spare parts, or complete spare systems/fixtures. That might mean a couple of spare already-programmed Arduinos. And of course, it helps to document everything.
Where I work we have LOTs of special test fixtures & test cables. I always try to get two fixtures/cables made so production doesn't shut-down while we try to repair one.
Over a long period of time, stuff can become outdated and impossible to maintain or repair. I've got some old test-computers and test-programs, and if/when they die, we'll have to start-over from scratch.
At home, I've got a couple of older microcontroller gizmos that I built about 20 years ago. That microcontroller would be hard to find and the programmer tools would be hard to dig-up, re-install, and re-learn. I'm not sure if I could find the schematics or software code. If I have to replace them, I'll start-over with an Arduino.
Hmap:
I have an idea for a small device to be used on a machine in a tier 2 Automotive Plant.The device will utilize two small steppers, read an analog signal, and communicate with a PLC.
In general do people use the Arduino products for hobby projects and then move on to something else for professional products, or is the Arduino product suitable for a professional finished device?
Thanks in advance for your comments and time.
Most Arduino boards use Atmel AVR chips. You should visit the Atmel site to see how serious they get, perhaps your automotive plant needs automotive-grade controllers able to stand more heat?
Arduino is a system based on the boards and the IDE. Do you care if the chip is source-coded in C or does it have to be assembler?
I would not recommend using a development board as your end product. The wires might come out of the header holes but for what you want I would recommend a Pro Mini with 168P chip on sale at DIP Micro for all of $2.36 and get yourself a development board (like Uno) and bits to prototype without soldering every step and change.
How is your C? Knowing C will really help you out.
Do you know "How To Make Many Things Work At The Same Time" and Finite State Machines?
We help people with those quite a bit.
DVDdoug:
Most solid state electronics is reliable[/b] as long as it's protected from electrical/environmental/mechanical stress.Over a long period of time, stuff can become outdated and impossible to maintain or repair.
I agree.
Industrial electronics has a lot of support. You have to simulate that to be trusted.
Build it strong. Industrial box and power supplies.
Build a second one as the parts are cheap. TEST IT. Run it and then swap it out for the spare and make the first one that works the spare so you know it will work.
DOCUMENT everything so it can be fixed easily. You will forget what you did, why you did it and where you got it. Write it down, leave a copy in the box.
Atmel isn't a toy company or some rinky dink science project operation.
They test and spec their own parts and have the docs to prove it including ISO certs.
Checking out any particular board source, that's a different matter. Don't shop blind.
You can build your own board if need be. It probably doesn't but you can.
What I like most about the AVR line is that the chip can run with only power, ground, and a bypass cap at 8 MHz on less than 3V (with reset disabled) and you can lock them to where it'd take a lab to read the flash. Everything you add is up to you, not required just to run.
Here we mostly deal with 8-bit ATmegas but do cover 32-bit ARM and others.
For your project, a $5 each Nano is probably overkill but allows you to develop directly on a breadboard with only a PC and USB cable for power, programming and debug I/O. The same Nano can be taken off the breadboard and wired with DuPont Cable female ends slid onto the header pins.
Modules to recognize color and switch power you can get off the shelf with pins to take the other ends of the DuPont Cable strands. You might need some voltage leveling but modules usually handle that too.
Arduino is a system that makes it easier for hobbyists and pros to use many of these controllers.
You can always buy the full pro Atmel programmers and software. It's just a bit more learning curve and budget is all.
first off, "Arduino" is both a piece of hardware and also a concept.
the concept is free programming software, a simple development board and a bootloader to take all the hard work of programing out of the project, limit the ability to interface with the microcontroller, but offer a very user friendly experience.
why an UNO or MEGA is not suitable ? there is zero protection on a pin.
24 volts, fried
short to ground, possibly fried
voltage spike, fried
emf interference, bad signals.
the list can go on. and there is a simple solution to each problem, and every poster so far on this thread can add another way to damage the pin, create a false reading or some other issues.
you can use an opto on an input to isolate the input from factor floor.
you can enclose the unit in a faraday shield, simple copper paint inside the plastic enclosure does wonders.
the point is, when developing on your bench, you only expose some risks, and even then, most of us understand what magic smoke is all about. on a very personal level.
if you look deep in your Allen Bradley PLC, you might find a chip that is just as fragile as the Arduino's 328, but you will find multiple layers of protection from the shop floor.
there is no reason you cannot use a simple board. as GoforSmoke said, a development board is great for the bench, bad choice for installations. pushing a wire into a friction slot is very risky, a compression slot that scrapes the wire is not much better although a little more secure, and a screw connection is vastly superior.
I like the nano plug-in screw terminal boards, but the plug in part is less than idea for the factor floor.
solder is the best connection, as are gold plated pins vs the lead ones.
rising cage screw terminals offer better connections than a screw twisting on a wire. the screw can displace the copper and twists to cut and deform the wire.
we can comment on almost every aspect of the physical board. then there is a the software, debounce vs. no. changing state from a 2.5 v input signal on a 5v digital pin....
There is no doubt that every single problem has been found, dealt with and available in some form or another.
there is an Arduino PLC, http://www.industrialshields.com/en/
This is not to tell you no, but to let you know it CAN be done !
that $1,200 PLC has basically, the same heart as your $2.31 micro.
just has layers of protection.
and, if your project is dedicated to one thing, you can protect for the problems of that one thing and not have to be prepared for every conceivable mishap that a PLC has to address.
Dave_vo:
DOCUMENT everything so it can be fixed easily. You will forget what you did, why you did it and where you got it. Write it down, leave a copy in the box.
brings me back.....
trying to sell some parts to an oil company in Saudi Arabia, for a laboratory.
document is not even close.
part numbers, values, 2 alternates, and suppliers contact information for resistors, solder, terminals, chips, LED's even the PCB itself.
way more than any normal project. when I asked why so much detail ?
"today the US is friendly and will supply parts. tomorrow, is uncertain "
Hmap:
I have an idea for a small device to be used on a machine in a tier 2 Automotive Plant.The device will utilize two small steppers, read an analog signal, and communicate with a PLC.
In general do people use the Arduino products for hobby projects and then move on to something else for professional products, or is the Arduino product suitable for a professional finished device?
in general, the chip manufactures make the same chip for hobby applications
then beef it up for a wider temperature for 'industrial' applications
then add protection for voltage swings and noise for automobile applications.
then test is 18 ways to Sunday for medical applications
then harden it for combat military applications. mil spec is not much more than industrial.
as you can imagine, cost for each level increases. and I would not be surprised if there was a $200 version of the 328 chip used in an Arduino UNO board.
what one does is start with the development environment and work out the bugs, verify that the chip will work for the application, then beef up the packaging, the circuits, the enclose and use a chip that is designed for the final application.
that is to say, stay with one chip manufacturer and one product line all the way through.
changing anything might mean starting from scratch.
only you can determine if the solution you have for the application will meet all needs and if failure will carry penalties.
we all go through a risk / reward procedure.
how many will I be able to sell if I saturate the market ? at one time the CEO of IBM said that the planetary market for a computer was , get this, he projected.... 5
"I think there is a world market for maybe five computers."
Thomas Watson, president of IBM, 1943
your project has some market, if you saturate the market in every tier 2 automotive plant in the US, will the Chinese/Taiwanese,/Japanese, out do you aren sell a better or cheaper version ?
can you sell it to the sensor manufacturer, or will the sensor manufacture make their own ?
will they add it to their devices like Microsoft did with all the software applications ?.
Just getting clean power can be a challenge as I recall from long ago.
We got a new CNC lathe in 81 and ended up with a huge transformer to filter spikes on the industrial park's lines. Hitachi wanted to keep that sale. 
Wow !
I am impressed with this Forum.
Thanks guys! (people)
There's also the Industruino
dave-in-nj:
and I would not be surprised if there was a $200 version of the 328 chip used in an Arduino UNO board.
Try asking for radiation-hardened, for outer space usage. I looked into it once. You can get them. The chips are $1000 each and there's a minimum order size of 8000 units. That's why some of the cheaper Mars missions are using the processor out of a Playstation - or something similar - NASA just buys 100 of them and tests the crap out of them. (No, they don't buy Playstations, just the chips.)
Thomas Watson was talking about computers that we would now call supercomputers. Computers so large that only a global superpower government would want to own one. Just a couple of years prior to him making that statement, a "computer" was a person with a mechanical calculator and the "computer department" had these people sitting at desks in rows.
MorganS:
Thomas Watson was talking about computers that we would now call supercomputers. Computers so large that only a global superpower government would want to own one. Just a couple of years prior to him making that statement, a "computer" was a person with a mechanical calculator and the "computer department" had these people sitting at desks in rows.
yeah..... cannot see that.... in 1945 an Arduino would have been considered a super computer.
and a lot is your vantage point. we really are getting close to that 5 computer mark.
there is the internet with hundreds of millions of terminals.
the NSA has one in the US, I think the military has one.
I suspect both China and Russia have one each for their countries.