4 ms·

Me neither. In industry, Arduino is almost an expletive- mostly associated with projects where someone is trying to use the hardware in a way that's not actuall

by fest 4y ago

Me neither. In industry, Arduino is almost an expletive- mostly associated with projects where someone is trying to use the hardware in a way that's not actually fulfilling an underspecified set of requirements, or where someone has stepped on one of many foot-guns in it's software ecosystem. An example for the first group: trying to save a couple hundred USD/EUR and using it where a PLC would be universally better choice (IO protection, EMC assurances, maintenance). Examples of the second group: delay functions disabling interrupts leading to missed serial data, (L)GPL'd library code, lack of debugging support (until recently), any sleep mode support. Don't get me wrong- there is a lot of good stuff in there- some of the libraries and projects built around it are really useful but effective use of them requires deeper knowledge than getting an intern "just doing it on an Arduino".

alar44 4y ago
I was offered a controls gig to design a very complicated machine that was to be used by consumers in shopping malls. Many moving parts. Many pinch zones, safety concerns etc. All in all probably 100 drives and maybe double as many sensors. They were hell bent on using an Arduino to "save money". I ran as fast as I could.
- iostream24 4y ago
  Clearly that calls for a 6502. Maybe do it in discrete logic wirewrapped
- fest 4y ago
  I've heard a story where somebody tried to automate some woodworking equipment with Arduino, bare optical encoders and microswitches instead of more rugged sensors. Predictably, the machine was constantly broken (a loose solder joint here, a spurious reset when some other machine was turned on there). Luckily,it did not lead to serious accidents, but eventually all of that was scrapped after wasting a few months and a controls company replaced it with appropriate solution.
  - alar44 4y ago
    Yeah this is what I was trying to explain. PLCs don't exist just to be expensive. They wanted to make hundreds of these and I was to develop a prototype. They also wanted it to be self-servicable. I told them they'd need to employ a dozen guys just to calibrate and maintain them.
duskwuff 4y ago
> using it where a PLC would be universally better choice (IO protection, EMC assurances, maintenance) Oh, that touches on another interesting and important point. Most development hardware, including Arduino boards, gets to skip a lot of standard EMI/EMC compliance testing because they're considered subassemblies and/or test equipment. Integrating them into a finished product means the integrator is going to be responsible for doing the testing and making any changes required for compliance -- at which point, they're probably better off building something equivalent themselves.
- iostream24 4y ago
  Which makes sense, given my gigawatt homebrew induction spike generators emf profile, seeing as how I only use the arduino to switch it to stun.
- fest 4y ago
  Yeah, and once you solve the issue of EMI/EMC compliance for generic applications (that is, expanding your market appeal to be large enough to sustain a hockeystick growth VCs expect)- you basically end up with a PLC, which means you now get the opportunity to compete with Siemens, ABB, Allen Bradley and the likes.

by fest 4y ago

alar44 4y ago

I was offered a controls gig to design a very complicated machine that was to be used by consumers in shopping malls. Many moving parts. Many pinch zones, safety concerns etc. All in all probably 100 drives and maybe double as many sensors. They were hell bent on using an Arduino to "save money". I ran as fast as I could.

iostream24 4y ago
Clearly that calls for a 6502. Maybe do it in discrete logic wirewrapped
fest 4y ago
I've heard a story where somebody tried to automate some woodworking equipment with Arduino, bare optical encoders and microswitches instead of more rugged sensors. Predictably, the machine was constantly broken (a loose solder joint here, a spurious reset when some other machine was turned on there). Luckily,it did not lead to serious accidents, but eventually all of that was scrapped after wasting a few months and a controls company replaced it with appropriate solution.
- alar44 4y ago
  Yeah this is what I was trying to explain. PLCs don't exist just to be expensive. They wanted to make hundreds of these and I was to develop a prototype. They also wanted it to be self-servicable. I told them they'd need to employ a dozen guys just to calibrate and maintain them.

duskwuff 4y ago

> using it where a PLC would be universally better choice (IO protection, EMC assurances, maintenance) Oh, that touches on another interesting and important point. Most development hardware, including Arduino boards, gets to skip a lot of standard EMI/EMC compliance testing because they're considered subassemblies and/or test equipment. Integrating them into a finished product means the integrator is going to be responsible for doing the testing and making any changes required for compliance -- at which point, they're probably better off building something equivalent themselves.

iostream24 4y ago
Which makes sense, given my gigawatt homebrew induction spike generators emf profile, seeing as how I only use the arduino to switch it to stun.
fest 4y ago
Yeah, and once you solve the issue of EMI/EMC compliance for generic applications (that is, expanding your market appeal to be large enough to sustain a hockeystick growth VCs expect)- you basically end up with a PLC, which means you now get the opportunity to compete with Siemens, ABB, Allen Bradley and the likes.