Spaghetti Detection Big Time Failure

AI monitoring was set to medium and running 1.04 firmware

Your eyes don’t monitor the first layer?

My Speg detection is not working at all.

I didn’t start this print. But really that’s besides the point, isn’t it?

It’s pretty common practice to monitor the first layer, at a bare minimum the first couple lines of a print. Even my 8 year old does this. Use actual intelligent monitoring instead of artificial intelligence monitoring.

Filament is also black and tough for the camera to pick up…

Yes I think that’s the core issue with the ai monitoring, but this print was a fail from the very first line and should have been caught by the operator

1 Like

It’s common practice to monitor the first layer because it’s common to have zero automated print failure detection on a printer.

I have been printing for more than a decade. Watching first layers is a well greased groove. This printer is marketed as having functional print failure detection. Why would an expectation of that functionality be weird to you?

It once paused my print (also black filament) because it saw a tiny piece of spaghetti, the size of a pea, but yet it somehow missed 30 minutes of failed garbage. I’m reporting this failure, not asking for opinions about whether you should babysit AI features or not.

1 Like

if you’re trying to report something for Bambu employees to see you should probably put it in the official bug reporting thread and follow the prescribed bug-reporting format, instead of posting a video in the general X1 forum.

I submitted an official bug report, via a support ticket, with all relevant details and logs.

I am sharing an experience of one of the X1C’s main features failing pretty badly, in the X1C forum.

You seem to take issue with this and insist on derailing the thread.

2 Likes

This is not constructive.

1 Like

sure it is. OP posted a video showing a problem with the printer, but the core problem could have been easily prevented in the first place.

Yup, and we could sit there and watch nozzle temp to ensure against thermal runaway. We can watch the spool of filament to ensure it doesn’t runout. We can do manual filament changes to make sure the AMS doesn’t mess up. We can manually shim the bed because we don’t trust mesh leveling. We can blow on the print so a part cooling fan failure doesn’t get us.

Or you could could realize that as technologies change, best practices change with them. The “core problem” here is that a main feature of the printer failed. End of story. Any upstream possibility of seeing the problem and intervening before the automated feature does, is totally beside the point and so obvious that it goes without saying. You seem to think you are bestowing wisdom here, when in reality you are just stating the obvious over and over.

1 Like

Its actually not, because the spaghetti detection is clearly failing here. You could tell someone that the airbag failing to go off in a crash could be avoided by not crashing, but you can’t say that the airbag is working.

I think you are looking for arguments and driving up the noise floor for people in this forum.

I don’t know if BambuLab has time for all messages in that forum (I guess not), but the new firmware update suggested me to turn all AI features on (because I had the bed detection off - I like using engineering bed with glue and 55 degrees C for PLA) and I didn’t agree.
I hope this is not what caused the next issue I discovered - false spaghetti detection. I left one of the print jobs 1/3 before its end. When I came back a few hours later, there was a message of “possible spaghetti detected” and the print job paused - the bed was kept warm for a couple of hours.
The stupidest thing is it stopped literally 5 secs before the end.
Now that I prefer to look at the print from time to time instead of wasting time.
Now that I turned the first layer inspection off too. using real intelligence :slight_smile:

1 Like

you got Cyprian working overtime removing posts hahaha

No I’m removing them not Ciprian. Your comments are somewhat rude. And user is correct, we should be able to detect failures like this. I am curious if first layer scan is on or off. I understand spaghetti is turned on, but curious about first layer.

As for watching first layer, I rarely do anymore, and I’ve started multiple prints from not even in the same room! So yes, our AI features should be working. User said they opened a support ticket, and yes we will address it via the ticket. There is nothing wrong with users posting their failures, and it is not easily preventable when the user may not even be in the same room.

Now Please Hobo be more polite ad less harassing of other users.

just ban me. I am who I am

First layer inspection was off for this print, and I imagine could have caught the issue. But, it’s obviously possible to have a “good” first layer come loose and give a similar result which only AI failure detection could catch.

Like I mentioned, I didn’t start this print. Someone else in my office did. I advised them to either watch the first layer or enable first layer inspection at the least. I think it’s somewhat annoying that first layer inspection is a global printer setting and not a selectable option at the time of launching the print. When you start a print, you know if you will be able to watch the whole first layer, and that is the time to decide.