Same, I left in 2014. I left after a couple of years to almost double my pay and not work on a Linux RT USB driver directly out of college for which I had no desire to be a SME, but in hindsight I think the coworkers I had there were smarter than anywhere else I've worked since (including Google). They paid absolutely nothing but seemed to have good culture, at least where I was at the time.
This matches my experience too. I'm a software dev in my 40s, started career at NI out of college, multiple companies since then.
NI had a fantastic college recruiting process. They sent engineers (not HR people or recruiters) to college career fairs all across the country, often back to their own alma mater, to snag the best-and-brightest from good engineering programs. They then organized big interview days where something like 50 or 100 college applicants would be flown in to interview on Thurs/Friday and then be taken out to downtown Austin on Friday night.
Their interview process was technical, rigorous but fair, and avoided the pitfalls of pointless puzzle questions or of favoring elite credentials over abilities. And their managers, product and sales positions all started from that same pool of entry-level engineers, so even if you were interacting with someone who's title was like "Product Marketing Manager" they had a BSc in engineering, comp sci, physics, math, etc.
The end result was that you always had very intelligent and capable coworkers surrounding you every day. I left for the same reason many other posters have mentioned- by mid career, I was getting paid substantially less than I could make elsewhere.
I had similarly painful experiences. Also the project I was on was mostly in maintenance mode, with mostly junior engineers out of college on it who had little experience with that particular code. Not many tests. Lots of code archaeology and "I'm crossing my fingers I didn't just break some subset of the enormous configuration space of this tool"
When I was there in the early 2010s, I met a greybeard hardware engineer with a big chip mask poster on his cube wall. He told me they lost the schematics and all documentation to one of their first timing chips that was used in lots of products. They ended up calling the manufacturer and asking for the mask image back, and any question anyone had had to be reverse engineered from the mask.
> Mostly hiring devs fresh from university so there was a ton of group think due to a lack of new ideas
This is a very interesting description of group think between less experienced developers. I've never thought of it this way, but it seems to have some truth in it.
That goes for every business. Their purpose is to increase shareholder value, not employee wages. The last part only happens due to market dynamics forcing them to, but they'd gladly pay engineers minimum wage if they could.
A lot of their products are relatively far into the long tail of hardware projects.
Enough demand in the market that it makes sense for a company to design and build them, but not so many units sold that all those design costs are easily amortized away.
I suppose it's fair to describe both as niche, even though basically every facility that requires high reliability controls (talking chemicals, energy, medical, etc) uses Emerson. They're enormous, but they do focus on one kind of thing.
It happens to also be what NI does (did) except NI did it in a way that was more accessible to education and hobbyists. Still expensive, but with things like educational toolkits using LabView as a base they have products that address the market for lower cost, lower reliability, but more flexible prototyping tools that I've never seen Emerson focus on at all.
But yeah, I knew about this a year ago because it's the kind of thing that matters a lot for my work. And since then I've known to not build any new prototypes that use NI software and instead move towards anything else...
> And since then I've known to not build any new software that uses NI software[sic] and instead move towards anything else...
Do you mean "not to build any new software that uses NI hardware", or are you specifically averse to Emerson's software dev practices even though you trust them to produce good hardware?
On that note, who is your new preferred vendor for DAQ hardware? Some of the stuff that NI allowed you to build with cRIO or PC-based multifunction DAQ hardware like their PCIe-6321 etc. was pretty unique. There's not a lot of off-the-shelf gear for on the order of $1000 that can do 100 kHz digital/analog signal acquisition.
I like Delta Tau PMAC gear for electronic servo control (though their recent acquisition by Omron seems to be having a similar impact as I expect Emerson is having at NI) and Delta Motion for hydraulics (not yet bought out by anyone, they seem to successfully transition to employee-owned after Natchwey retired last year, but time will tell)... but neither is a true multifunction DAQ system like NI.
At my company we’ve had success with Advantech (Japanese) DAQs.
I always considered NI best for research labs / benchtop measurements, and this acquisition seems like a poor fit because NI was pricy but with good end user support, while Emerson likes to have very expensive/little support.
Well, I'd say that NI hardware was at one point the best you could get in terms of performance and ease of use, if you had tons of money. Quick prototypability is the main advantage compared to everything else.
I'm not sure who I would use for that purpose, I'd say that we're blessed with a huge variety of things which are of quite good enough performance and ease of use, but also with the curse of needing to deal with different vendors for different kinds of uses. I wouldn't expect most companies that sell general purpose DAQs to also sell FPGA systems for signal processing, for instance. That's odd.
But if you just mean digital and analog inputs and outputs there are so many options now that are adequate, even just an eval board for a microcontroller running a LabView firmware module can do well enough (and this route lets you escape NI better later anyway). This is a kind of cool new thing, using an Arduino or whatever as if it were an NI DAQ.
So the ecosystem is much larger, and less locked into NI software (LabView anyway).
On the other end where reliability is important, I wouldn't trust many places, but I barely trust NI. Honestly I'd be looking at Emerson if I wanted that level of assurance, but I wouldn't use that for prototyping. In a prototype I have to design it to fail safely enough rather than trying to make it foolproof.
So for prototyping there's a ton of options now, and once the prototyping is done -- if it needs to be high reliability then it's probably going Emerson or something similarly expensive anyway, and if it needs to be low cost then you'll just put it on an embedded DAC and ADC with a ARM-M0 or RISC-V or Raspberry Pi or whatever you fancy. And if you want it to be more expensive but more reliable you buy some kind of OEM module you put on your PCB so that they can be more carefully integrated while using certified analog input and output modules.
I think that ends up being three widely varying use cases, and prototyping is maybe just too easy to do in other non-LabView and non-NI-hardware systems such that the high-reliability Emerson-style market is what remains. I won't miss LabView, that's for sure.
I must have missed this when it happened last year (April apparently). Perhaps not surprisingly I associate the name "Emerson" with comically large integrated stereos (boom boxes). But apparently they are building up a test and measurement group.
I am aware, and it is unfortunate for Emerson Electric in terms of branding things in the electronics space. I don't doubt they will retain the National Instruments and/or Digilent brands for that reason.
It will make me chuckle every time I see that Digilent banner "An Emerson company."
They were bought by a private equity company in 2021. Would love to know more about the Berlin programming scene in the 90s. The initial minds behind Generator/Reaktor were incredibly inspiring.
