Why do we have flow rate calibration, resonance calibration, lidar & AI and still have to fine tune EVERY SINGLE PRINT? Like cooling/heating, fan speed, print speed, over/under extrusion, belt tension, etc. Why do we need to tweak all of these?
IMHO, all we need is color settings and quality settings, like
AI driven HW & SW should do all the stuff for us.
Am I right?
All comments are very appreciated.
Because stuff happens and automatic systems don’t always work 
AI should fix the stuff from happening.
Not likely, liveware will always have the upper hand where messing up is concerned.
You don’t if you always use the same filament and print the same model. 
Boring… Imagine 68.589th benchy coming out from your printer.
I’ve never once printed a Benchy. Just to let you know, most everything changes from one day to the next–Like weather, temperature, models, speeds and filament. I rarely have to change my filament parameters. I usually shoot from the hip on new filaments after considering the seller specs.
OH yeah, there’s also the user errors. Like dirty plates or wet filament.
Dirty plates: lidar => err msg
Wet filament: flow rate calibration => err msg
P.S. I have printed only one benchy so far… first one after 4 years.
Preface:
Your eye processes in 24 hours the amount of data from a 90-year-long HD movie - so it’s all a question of what attracts your attention. But you’re missing a lot more than you ever can realize. By the way: smeels, tast, skin feelings and so on are additional infos which are still not includet into this 90 years of HD watched in 24 hours - different people have different weightings but overall everyone perceives about the same amount. So if you catch something, you’re going to miss something else that other people perceive.
Your case:
You scan a very small part with the very precise organ “eye” that is made to constantly scans the entire world and the AI ​​can’t follow you yet. And by the way, I would never call anything in a printer AI, because AI trains itself. At least from my point of view, that would be the most elementary basis for any existing intelligence. To be able to train yourself like you train your eye day after day and get better and better at recognizing the smallest errors. You can see this also well by industrial visual inspectors - they are getting better and better and can see the smallest deviation more and more precisely day by day becouse they get more and more trained by themself and more practice…
My case:
AI - It is enough that I no longer have to look for function-oriented parts or wait until the first layer is printed or it is dressed and I notice that after 3 days the 10 hrs. Printing is not yet finished. Sure I can do it better, but it’s better than if I never look at.
From my point of view, as a programmer I recognize AI as algorithms and data. AI cannot train itself since AI has no idea what training is. It’s just an application. Very smart application… but it cannot think itself.
Many printers are now connected to the Internet. They can send and recieve data to and from appropriate databases. They can enhance the knowledge base. And humans can improve AI algorithms so that machines can get “smarter”.
Maybe some people will have more to say about AI. But, basically, engineers should be able to implement smart algorithms anywhere they need.
And I am 100% convinced that engineers and programmers can do this. So can the engineers and programmers from Bambu Lab.
Well, the main Swiss national television channel was on this topic for a whole week - I didn’t watch everything but it was interesting - and of course i fare of to be informed about. Of course I couldn’t see everything since it runs one week thrue every topic. But what I see was impressive.
Of course an A.I. doesn’t know what training is (Maybe? I`m not an expert), but an A.I. can train herself. The oldes example chess games against herself. And now in a very short space of time, more than a human could ever play and gets better and better - found possibilities that no one recognized and now even strategies that people have adopted. Or a more modern example, a robot dog that prevents one fall after another and gets better and better by himself. Of course, they also went through topics like deep facks, medical recognition and so on, and it was really on air for whole week. Also like quantum processors, a new center has opened up in Basel Switzerland, which analyzes paths as a whole and no longer calculates them route by route and getting her own way by herself to do so…
It was also interesting that you saw 3 different A.I. in a Quis show “1 against 100” - where one candidate competed against a hundred people in questions with jokers and so on - an A.I. won, the other lost by the first question - the trainer of the AI ​​also showed very well during the show how he can unsettle her, how she questions herself and how she leads herself on the right path.
And by the way, at least in German, Inteligance stands for able to think (simply put), so if it can’t think - then the I in KI is used incorrectly and it’s just K
Es gibt keine KI es Gibt nur Menschen die Programmieren und im schlimmsten fall sich einbilden es wäre KI
Also es ist nur Sensorik die check und ne Korrektur wären einen Druckes macht die if Else Anweisung + Möglichkeiten also While schleife oder andere las das Decken es bringt nichts (-;
They are options on by default that we have the ability to disable (uncheck and will remain off until rechecked) to help those new to printing. As they gain experience they can disable them and move on. I printed for a few years on Prusa and just switched to Bambu due to it being locally available and pricing being insanely good.
I have recently uncheck Flow Dynamics at each print but I do a calibration and test run before a big project.