Hi, I'm new here


#1

Just wanted to introduce myself. I actually work at one of the auotmotive OEMs in the Southeast Michigan area, in the United States.

Spent basically my “birthday weekend” playing this 'til wee in the morning.

This is a very good game. Going to try out more stuff when I have more time.

It would be interesting to see how this game compares to some of our actual final assembly plants, but I am going to try some different ideas in the sandbox/freeplay mode.


#2

Welcome aboard! I hope your stay here is an enjoyable one. :grinning:


#3

Thank you.

Yeah, I stumbled upon this game somewhere and bought it and started playing it.

As someone who works at one of the automotive OEMs in the Southeast Michigan area, this is very intriguing and fascinating to me.

A definite shout-out to @cliffski for creating this game.

I’ve probably been playing it now for between 48-60 hours and there are definitely some bugs, but overall, the game works well.

I have a massive factory that I am about to revamp tonight because I am running into bottlenecks so I am about to start reverse engineering the architecture of the game in order to optimise my production.

It’ll be interesting to study, for example, the difference between the Ford Production System vs. the Toyota Production System.


#4

Welcome! Feel free to post any bugs you discover, I fix them as I find them, so its likely any you encounter are things I do not know about yet :smiley:


#5

@cliffski
Thank you.

Yeah, like I said, I started playing with it and then got REALLY interested in it (as I mentioned, I spent my birthday weekend recently basically playing your game so that’s a LOT). Shhh…wife doesn’t know. She thinks that I was “working”. :slight_smile:

I had modified the giant factory map so that it would let me build a little bit more freely because real final assembly plants can be 8 km long in real life.

I’ll file an official bug report later (but I wanted to read and review the existing bug reports first), but like two things that I have already discovered are:

  1. If you have local production of components (brakes for example), and you place it in one corner of a large factory (to keep the line and the in-house component manufacturing and support separate), the game currently might complain about the “fit brake” station being out of resources even though there are nearby supply stockpiles (filled with brakes) and that I have something like 278 brakes currently stockpiled according to the component screen.

It doesn’t use it. Or it doesn’t know that it’s there because where it is manufactured/stockpiled is too far from where the brakes are being installed/fitted.

So that’s a bit of a bug. (I’m currently playing version alpha 1.60).

  1. The other bigger “bug” is the lack of sequencing for the production. So, when you fit the front axle for example, you can set and define the different models and trim series of vehicles that you are going to build (which in your videos, you call it Sedan E and Sedan M for example). However, for example, when it gets to the “fit brakes” station, if a vehicle has ABS, the station doesn’t know about that in advance, and therefore; will issue an “insufficient resource” warning/error until the brakes are installed and there is room at the “fit brakes” local stockpile to pick up 4 servos for the ABS. In my case, my manufacturing facilities for the servos are all the way at the other end of the factory and even though I have local stockpiles of it, again, it doesn’t appear that the “fit brakes” stations know about it or isn’t “smart enough” to search wider/deeper.

Again, since the game actually does a FANTASTIC job of mimicking the real world production environment, I can also tell you that in real life, the production itself is all sequenced so that the right parts, the right AMOUNT of parts, will be at the station at all times.

This element/aspect is currently completely absent from the game entirely, so it makes it impossible to set up even a relative simple build schedule of just four different sedan trims, where some have ABS, some don’t, and the various combinations of other options that are available.

Again, I’ll file that as a “bug/suggestion” officially later on, but these are my observations so far (when I play it in freeplay mode) so that I can actually try and set up a grossly simplified version of a style of final assembly plant that we have.

The last thing that I wanted to mention/ask you directly is if you build ONLY electric vehicles (i.e. no vehicles that uses petrol), do you still need to have a “fit exhaust” station?

For example, if you actually want to try and set up the Tesla Model S factory in your game, do you still need the “fit exhaust” station? If it doesn’t, GREAT! If it does, then that probably doesn’t make sense, huh?

(I haven’t actually tried re-creating the Fremont Final Assembly yet, but that’s my next “pet” project in your game.)

Thank you.


#6

Interesting stuff. its amusing to think our big factories are too small :smiley:
The point about the production being too far away does sound like a bug. It should not matter at all how far away stuff is if you tell the slot to prefer locally made parts. That sounds like maybe the routeing table has not been updated properly somehow. I shall investigate.

The second point, I am well aware of, but actually its almost ((impossible to fix.
The system in real-world factories works because:
#1 You know which car is coming next because there are single lines without multiple T junctions leading into them
#2 The line is not broken up or rearranged while cars are on it
#3 The specs of a car are never changed when it is inside the factory.

Sadly all 3 of these things can be initiated by the player, meaning that we absolutely can NOT know which car will enter our slot next.

Now to a player (rather than a CPU), a lot of the time it looks stunningly obvious which car is next, and what it requires, in cases where there is a single long conveyor belt with no branches, but the game cannot rely on that. In theory, it could scan ahead 3 or 4 slots, and in cases where there was no T junctions, make guesses based on that, but even then, the player could change the design of that car the second before it enters the slot…

In practice, the player can avoid too much hassle in that area by simply ensuring there are production runs where for example 50 sedan Ms get produced, THEN 50 sedan Es, and so on… This lowers the design changeover time.

And yes, electric cars can skip the exhaust and radiator sections. My current factory uses a smart junction to sort the electric designs so they skip right past them :smiley: They still go to fit fuel tank, but gets batteries installed there instead.


#7

@cliffski
Thank you for your response and reply.

It’s great that you are willing to take the feedback.

Yeah, I didn’t know what the range of search is when it goes to “look” for resources.

The factory that I have been playing actually became perhaps too large for it to load (submitted another bug report for that) because the .xml file alone is approximate 5 MB, so it spends quite a lot of time processing it when it tries to load it. (I wanted to be able to send you a copy of that xml file so that you can play around with it to see where problem areas are with the game - and admittedly, it’s not the best, most optimal factory design yet, because I am still in the process of reverse engineering the game engine in order to come to the optimal factory design, but I couldn’t upload it here. So if there is a way for me to email you the xml file, just please let me know. Thanks.)

(I’m playing this on an Intel Core i7-3930K, 6-cores, hyperthreading disabled, with 64 GB of RAM, and a Quadro 5000, I think.)

re: sequencing
Yeah, I get that. In fact, because my large factory yesterday didn’t quite load, so I switched over to building a new one because I was trying out something else in order to make the production go faster/be more efficient. (I found a rather quick way of building at least the basic sedan).

And your points are very valid. What’s interesting is that in the real plants, there aren’t technically “T-junctions” for the supply conveyor system, but not all real world final assembly plants also have supply conveyors either. A lot of the material handling is actually done with “hi-lows” (I don’t know what they’re called in the UK). A lot of the plants also have robotic/automated supply delivery vehicles that run on the plant floor (and also have right of way on the plant floor because it has limited sensing and object collision/detection/avoidance capabilities). So to that end, real plants actually have an infinite number of T-junctions only because there will be as many T-junctions on the plant floor as there would be for any other traffic (pedestrian and/or motorized (or electrified (battery powered)) vehicles).

I realise that for a game, incorporating this form of sequencing can be MASSIVELY complicated because where you are absolutely 100% correct/true is that the assembly conveyor usually will run linearly, unlike this game, where you can have it run in parallel.

re: EVs
I just wasn’t sure what “slots” (i.e. stations) pure EVs were able to skip vs. what’s absolutely required. Thanks! I might still experiment with trying to create like the Tesla factory later. I’m still trying to reverse engineer some of the core elements of the game in order to drive production up. (I’m currently importer-limited to about 49 vehicles per hour, so I’m trying to offset that/provide some relief by installing and introducing local, in-house production into my factory.

Again, thank you for creating this game. I love it!


#8

Yup the whole conveyor belt system in the game is a bit of a nod to how people outside the industry think cars are made, compared to the modern reality, which as you say, seems to involve autonomous skates and other vehicles moving cars and parts around the line.
Having said that, the pickup truck prod line I visited in Michigan still looked very much conveyor-belt driven, and had a huge amount of manual labour, although I suspect that says more about Ford than the modern industry :smiley:


#9

Yeah, then that’s probably the River Rouge facility that you visited.

So…I’ll be honest - I’ve never actually done the public tour, but I’ve been through that plant, getting to other areas that aren’t open to the general public, on official, company related business.

Having said that, modern assembly plants (at least as I recall from my detailed walk through of GM’s Lansing-Grand River Assembly plant (which used to make the Cadillac STSes, but I don’t really know what they make nowadays) – a lot of them are structured like this:

  • There is a body shop area where the body-in-whites are assembled. Sometimes, everything is “dunked” into paint (well…that’s how it was like 15 years ago, now they probably spray), and so sometimes they would assemble the vehicle and the disassemble them for paint in order to make sure that everything fits first. Paint shops are also usually sealed in order to prevent contamination.

  • Then there are also usually two lines that run in parallel, but vertically (second floor vs. first floor). LGA had it where the chassis and underbody stuff would be assembled on the first floor while the second floor would assemble the body. Once they’re both built up, the conveyor would bring the body to be married with the chassis at which point the engine would basically slip in underneath the entire car and for example, again, if I recall correctly, they would drive the first 12 screws/bolts of the suspension strut into the body before anything else.

So in your game, you have those operations as being sequential to each other, but in many final assembly plants now, the reality is that many run in parallel. That’s probably what you were seeing with the conveyors. So, it’s not uncommon for the carrier (a.k.a. “clamshell”) to be carrying to body around for different operations (along with its associated parts). Thats probably what you saw.

Another example as I mentioned is that they might put the doors on the vehicle once to make sure it fits properly, and then take them back off for it to be painted (so that it will get exactly the same colour as the rest of the body) and then once it dries, might be sent so that the door could be built by adding the glass for the windows, motors, controls, switches, speakers, handle, latch, and door trim, for example, all separately before the door goes back to be mated back up with the original body that it was “paired” with. (Again, as I recall, LGA was set up that way.)

re: manual vs. automation
That’s actually usually quite the debate. A lot of people tend to think that many of the repetitive tasks (of which, there are countless), can be automated, but the reality is a lot more complex than that. Setting aside the human element for a moment, it wasn’t that long ago that Toyota and Mercedes, both replaced some of their automation by putting more people back into their respective processes probably because automation wasn’t helping them in those areas.

Elon Musk (Tesla) is famously having said that he was oversold on automation and he too, also had to hire more people in order to address the production problems that they’ve been having - the extent that he’ll talk about how he’s sleeping at the factory to get stuff done.

Musk fans thinks that’s nobility. As someone who’s been in the industry for a while, I think that it means more that he doesn’t know what he’s doing or talking about (because no other CEO has to PERSONALLY get involved like that which means that some of their processes are massively broken if the most expensive guy has to step in). But that’s just my take on it. (IF things work as they should, the CEO shouldn’t need to be at his own factory, fixing things.)

And the UAW would always want more people involved, so like so many things, there’s a time and place for it all.

The ground “belt” that you’re talking about - that’s also pretty much an industry standard as well, but to varying degrees. Some would have a lift so that it would raise or lower the height of the thing they’re working on in order to accommodate the worker, while other resolve this using alternative methods/means. So it depends. In-ground conveyor systems tend to be favoured, mostly due to their load carrying capacity. You’ll also note that some final assembly plants, where the workers stand, they’re also standing on a moving conveyor belt as well so that the car is never actually, truly stationary.

Some of my mates in university studied industrial engineer, and went on to become some of the engineers that would be part of the team that would design these plant systems. If you think engineering a car is difficult, try engineering the stuff that’s needed to build said car. It’s quite the sight to behold.

If you get a chance to visit as many plants as you can, you should take every opportunity you are presented if this is something that interests you.


#10

Yup I read a fair bit about the whole separate lines and ‘marriage’ thing but I thought that new players might have some problems with the concept of their being two lines which then have to meet. also from a visual POV, people seem to enjoy seeing a car progressively build up from completely nothing in stages.

I really should have had people working on the lines a lot earlier in the development of the game, even if it was just a few slots (perhaps some of the fiddle electronic fitting stuff) so the player could research automation in stages to replace people with robots. I still think this might be a good idea.

I read a lot about teslas automation and backtracking, but I don’t think they have entirely given up on that. They seem to be determined to use a lot more, especially in stuff like fitting seats etc. I heard rumours that the benefit of the glass roof in some tesla models was that it allowed robot arms to get inside the car at a later stage of production than would normally be possible, but that might just be an assumption.

Its all totally fascinating stuff. I must plan a proper factory visit again.


#11

No, I don’t think that Tesla has competely backtracked being that they have had so much invested in it by now. And that’s always the difficult part about when you make a business decision blunder of that scale, cost, and magnitude that by the time that you’ve realised something is wrong, it’s usually far too late in the process to completely reverse it, so you’re stuck with “bad” stuff (less than optimal designs, plant layout, etc.).

re: glass roof
Yes and no.

Of course, as it is with automotive engineering, there are always trade offs. So, you might be able to help you manufacturing with something like that, but then it also comes at the additional engineering cost because the vehicle itself still needs to be able to meet (and in most of Tesla’s cases, exceed, by a significant margin, the requirement), so in the end, you just pay for it with something else.

re: visiting factories
Yeah, I wished that I was in a position to be able to invite you to a visit to one of our plants for a more detailed, in-depth look.

Unfortunately, I don’t have that authority. But if you write to the company, maybe they’ll let you.

The most exposure to final assembly plants was when I used to work in the supplier base. There, I was able to visit many different plants acrossed different OEMs so that I can see how each OEM solve the same problem, but in different ways.

It’s even better when you are with someone who can explain stuff to you as you go up and down the line.

I think it’s fantastic that you wrote and developed this game which introduces people to these things that so many other people wouldn’t have or been able to experience or exposure to it.

Thank you.