Cloud outage causes Bambu 3D printers to start printing on their own
themessenger.com
themessenger.com
The printer has to be connected to the “cloud” in order to access the webcam, control the internal lights, and control the print head and heated bed from their Bambu app. However, part of the initial setup process requires you to connect to the printer via Bluetooth in order to setup the wifi connection in the first place! Clearly it IS possible to communicate with the printer without an internet connection.
They intentionally ruined the user experience in LAN-only mode to persuade people to connect their printers to the Bambu cloud service.
Things like the nearly required cloud integration were glossed over with "that's normal today and you get so many shiny features with it" and hardly anyone challenged Bambu by saying that most of the features enabled by cloud integration didn't really need the cloud to be implemented. Things like immature firmware were glossed over with comments like "there are some bugs but that can be fixed over time, so just go ahead and assume that this young company without a proven track record will provide good support to these printers."
Add the marketing to the consideration that there were no other pre-built "fast" printers when the Bambu printers hit the market. Your only other option was to build your own like a Voron, which many people didn't want to do. Now there are other pre-built options that don't require cloud integration, so people can really weigh the pros/cons of the other options, but Bambu got an advantage by being first to market.
When the Bambu came out, there was nothing like it on the market that combined price, speed, and ease-of-use. Yes, there were cheaper printers, and faster printers, and easier-to-use printers, but not in a single package, and especially not in a pre-built package.
Things like immature firmware were glossed over
The "immature" firmware is still better than the "mature" firmware Creality ships. And Prusa shipped buggy firmware for its multi-material addon for several years, despite being in the 3d printing game for over a decade.
Bambu got an advantage by being first to market.
Bambu's first printer came out last year. The Anker Make came out before it. Creality and Prusa have been shipping 3d printers for a decade. Voron has been a thing for about a decade. What is this "first to market" advantage you speak of?
Voron is a kit or diy. Creality printers that were available at the time were not fast like the Bambu. Their K1 came later.
I wrote that at the time Bambu was the first to have a pre-built fast 3d printer. That is the first to market (the pre-built fast printer market) that I was talking about.
I don’t know how you say I was wrong, because everything you “corrected” me was in my previous post.
are they hoping to monetise you in a future update?
continued use? $19.99/month
So it appears you're angry about a strawman that doesn't actually exist.
... yet
For instance, with my last 3d printer, I had to open it up and jump some pins on the mainboard, and then had to get a specific piece of (cheap and readily available) USB hardware to write to it from a laptop - and even then, it only worked from a specific USB port on some of my machines, because of weird USB host incompatibilities. And my 3d printer was one of the most common buy-this-one-first products of its time!
Oh, and there was also some weirdness with the PSU as well that somehow caused flakiness when powering up the webcam when I tried to share power for the pi with the printer itself. I'm comfortable with troubleshooting that, but I wouldn't expect nor even want the average person to have to bother with that kind of stuff.
So the appeal/marketing of the Bambu products is that it's not something for tinkerers - it's something that the average person can just setup and immediately get reliable prints, with all the bells and whistles on top of it.
In other words, the idea is that it (quite literally) "just works". Even with the drawbacks that we all know about proprietary devices and cloud services, it does its job for the average person.
The positive side of all this is that Bambu's relative success means a bit of pressure on the other leaders in the 3D printing space, and we can see that they're all iterating quickly and well to compete in aspects where they may have been resting on their laurels before.
I had to do that for almost every new game and hardware upgrade on my old 386. Welcome to the club, pal.
Most people just want a printer, not two new hobbies.
Checking the progress of your prints remotely and stopping them if something has gone wrong. There's options for spaghetti detection and so on, but those aren't perfect, and other kinds of failures can happen.
Just like most appliances, a 3D printer will only catch fire in the case of a serious hardware malfunction or egregious user error. The point is for this to be rare enough that the risk is negligible, so all you'd worry about is your print failing.
You can make it work over the internet without a cloud connection, but it's fiddly and requires set-up (port forwarding or hole punching). People don't want that, and Bambu's value proposition is zero-tweaking at high performance - otherwise you'd go for another model.
A pinkie promise that, "We'll do better after you got burned for our data-collection greed" falls a little flat.
The camera we’re talking about only sees the inside of the printer.
Devil's advocate, inadvertent peephole cameras and reflections in surprisingly small regions of interest are a fact of how light interacts with glass, metals, low-tolerance parts, and the other sorts of things that make up the printer. There are lots of recent fun papers inverting the pattern going around a curved painted corner or through a hole accommodating the string holding up your blinds or whatnot to get a detailed view of what's on the other side.
Their malicious use cases are probably less cutting edge.
This is actually a pretty bad argument. Bluetooth only goes 30 feet, so your connection would likewise only go… 30 feet. All of the things you can control on your smartphone can be controlled on the printer itself, so what’s the point?
Unless you are complaining that you can’t use the smartphone app, while at your own home on Wi-Fi, without connecting to the cloud. That is, I will just say right now, not how 90%+ of customers would expect an app to work. For most people, a smartphone app that only works at home is confusing and upsetting.
Finally, while it doesn’t work from the app, you can control those things and view footage using the Bambu Studio desktop app without a cloud connection, IIRC. You just need a PC, not a phone, though I would argue this makes sense as a fair trade-off considering the consumer demographics. A PC would be way more reliable anyway as your phone would be switching onto cell tower networks which obviously don’t have your printer on them.
Edit: Also, Bluetooth tops out under perfect conditions at 2.2 Megabits. Real world is far lower. You can’t reliably send a video stream over Bluetooth. Heck, Apple doesn’t even try to send a lossless audio stream over Bluetooth.
I may not have made myself clear with my original argument, I meant to say offline controls are clearly possible. So if offline controls are possible there’s 0 excuses for controls not to be available in LAN-only mode. (LAN-only is wifi connected, cloud disabled)
> All of the things you can control on your smartphone can be controlled on the printer itself, so what’s the point?
It is way easier to tap a button on the app when dealing with hardware issues than using the built in menus (that’s my opinion). But more importantly there is no valid reason to not make them available in LAN-only mode, that was an intentional decision on their part.
> That is, I will just say right now, not how 90%+ of customers would expect an app to work. For most people, a smartphone app that only works at home is confusing and upsetting.
The 3D printing community is pretty technical, so I think anyone disabling cloud mode would be able to understand the consequences. An opt-out (without degrading the experience) is all bambu owners are asking for. And keep in mind 3D printers are a fire hazard so printing while you aren’t home is a risk.
https://wiki.bambulab.com/en/software/bambu-studio/remote-co...
You just can’t do it from your phone. Which, considering your phone likes connecting to cell phone towers (and thus disconnecting from your printer), I completely get why they didn’t want to bother.
I don't know what kind of weird broken phones you have, but every smart phone I've ever used has had no trouble connecting to things on local networks via wifi even when not in a cell deadzone. Your objection sounds made up to me.
Now to address your original concern about Bluetooth. Bluetooth is not the best fit but Bamboo printers do have WiFi. They can host local web UI which is what pretty much every other product does. It requires will and resources to support offline tooling for users though. It also reduces the probability that the user will subscribe to their cloud services, if any.
Coming to the point of an App only working at home. Well, there are several ways to address this issue WITHOUT requiring a user to use their Cloud offering. A simple way is to EDUCATE your users that the app will only function on local wifi. Average users are smarter than you think. If the user would like their app to work when they're not home, offer them an option to connect to the cloud but do not make it mandatory. Next, the printer can use a combination of UPnP, Dynamic DNS to expose it over the internet. Yes, it is scary to do this but with the right security practices it is possible to expose your printer over the internet so your app can reach it. You don't need their cloud to be the intermediary.
Finally, these printers will likely stop working entirely if the company goes out of business or decides to wind down their 3D printing business or just plain planned obsolescence of specific models. All in all, subscription & online paid services for hardware products is an anti-consumer model.
> Finally, these printers will likely stop working entirely if the company goes out of business or decides to wind down their 3D printing business or just plain planned obsolescence of specific models
They can be set up without internet. They can print from a microSD card without internet. I have a Bambu X1, have never connected it to the internet, and everything works fine. No worse than a Prusa from 2 years ago would have.
Except, even better, I can connect it to a LAN, without internet, and access the camera from my PC using their open-source desktop app. I can’t do it from my phone, but my PC isn’t hopping between networks constantly.
Saying it will die if the company goes bankrupt is pure FUD.
I don’t know if it’s worth addressing the other points considering this level of ignorance. Especially this:
> Next, the printer can use a combination of UPnP, Dynamic DNS to expose it over the internet. Yes, it is scary to do this but with the right security practices it is possible to expose your printer over the internet so your app can reach it. You don't need their cloud to be the intermediary.
This has the reasonability of the now-infamous Hacker News diss when Dropbox was invented.
https://news.ycombinator.com/item?id=9224
Also, we are in an IP Address shortage. More internet companies like StarLink are disabling port forwarding because they don’t have enough addresses and the cost per address is currently insane. It might not even be an option for a growing number of people out of technical necessity. In addition, many ISPs, including mine (Xfinity), reserve the right to ban you for running a home server of any kind - Bambu obviously doesn’t view encouraging people to persue activities that could get them banned from their ISP as a viable solution. Even with a warning sticker.
In general anything that contains a heating element or that generates or uses power in the KW range should not be internet accessible. Too many griefers in this world.
So their current product offering includes a heating element capable of starting a fire but no monitoring or cutoff?
For example, your workshop may be well ventilated during the day and completely closed up at night. All the safety features in the world don't add up to a 100% safe usage experience for any piece of hardware. You don't want your printer doing this for the same reason you don't leave your table saw running when locking up for the night. Technically safe, but unadvisable.
Yes, obviously. But that doesn't mean you should omit basic safety features.
Extremely broad and reliable rule of thumb: hardware safety features > software safety features
The typical disaster on 3D printers is that the print detaches from the bed and sticks to the nozzle instead, which the printer doesn't notice, and so keeps extruding plastic into this obstruction, resulting in the nozzle being encased in a big blob of plastic.
This isn't immediately a fire risk, because the hottest point will still be the nozzle which is temperature-regulated, but due to ongoing mechanical damage you start to risk things like the thermistor and heater wires being damaged, and that can cause havoc.
Prusa recently introduced thermal modelling on their printers which detects heating anomalies like this, because the hotend becoming encased in plastic changes the rate of temperature leakage to the environment, and so the required heater power. The idea is that you can catch the issue at a very early stage.
The combination of big government and big tech has normalized so many destructive corporate behaviors, that I fear we will never be able to fix the problem from within the system, nor will a successful revolution ever happen. Humanity is absolutely fucked, if we continue on the current trajectory. And no nation that has any significant influence is without fault. The modern "civilized" world is garbage for the vast majority of individuals.
Cloud is for tracking.
I still wouldn't use somebody's else cloud for them, but they are very useful things.
Equivalent systems from other companies do have subscriptions though.
So the cloud is seen as a requirement, and in the end, no desktop app gets created.
I’m able to print perfectly well over the LAN, thank-you-very-much.
However, I say all this as someone that is technically inclined, wants to tinker, and is a fan of the open source. Somebody that wants plug-and-play should definitely look at another printer.
If you want to build your own printer for the experience, that's obviously worth something, but it doesn't make sense from a purely financial perspective.
I'm looking at pre-assembled printers to add additional printers to my fleet, because I'm not keen to repeat that again.
That said, the amount of really nice mods for Voron due to its open source design has been fantastic and I already consider several of those to be essential in a new printer...
What benefit does "cloud" offer a 3d printer? Couldn't all the same features be offered with local computing? Isn't "cloud" just a route for trouble for almost all users?
I'm leery of the whole "image recognition to spot errors" thing anyway. you've got far more direct inputs to monitor. is image recognition truly that reliable or is it not just a source of more errors when the recognition software goes awry?
> What benefit does "cloud" offer a 3d printer?
A large print can take 24 hours or more. Offering the ability to walk away and check-in and issue commands remotely is useful to people.
(Yes, you shouldn’t leave a printer unattended, but the basic reality is people do leave their printers unattended anyways.)
> I'm leery of the whole "image recognition to spot errors" thing anyway.
I don’t think it’s inconceivable that it could work well. The main issue being monitored for here is “has the print prematurely detached from the bed”, which I’d imagine should be possible to reliably detect with a known camera, angle, and print characteristics.
> I don’t think it’s inconceivable that it could work well.
It's been done before, although others with actual experience of Spaghetti Detective will have to chime in with whether it works well.
This isn't intended to spot subtle layer shifts or something, it's more about fatal errors that left unchecked will just throw away filament or (in a worst case) ruin a printer.
The Bambu systems work just fine from a microSD card without an internet connection at setup or in use. The Cloud is optional, and I think that any user who decides to use the cloud should be aware of the logical risks included.
> Couldn't all the same features be offered with local computing? Isn't "cloud" just a route for trouble for almost all users?
Many can, including viewing the cameras, which now work just fine with LAN mode on Bambu printers. However, I would argue people want to remotely check on their prints (what if I started a 12-hour print, left for work, and want to make sure it looks good at lunch?). The alternative there is opening ports on their router which, for the average person, is a much worse and less secure idea.
just make people a printer and you don't need to develop a massive IT overhead
wouldn't this give you the ability to sell a cheaper product while marketing yourself as "not selling your customers information"?
The reality is that Bambu makes a premium product, in part precisely due to the cloud features. Which you can turn off, or never activate, but no-one does.
Printers like these really genuinely do benefit from cloud connectivity. Unfortunately that doesn’t mean you can skip on QC…
but think of all of those off-shored jobs you'd be taking away!
1. Any personal images/data sent to cloud need a FDA approval. (FDA = Federal Data Administration) 2. Any person working in this agency is banned for life to work in the private data sector for life (it better be lucrative to work there). 3. ... perhaps fill any loopholes you see here as you move along ... (along with percentage of revenue punishments for each loophole used first time)
So not just new regulation, but an entirely new regulatory body. Also, the Food and Drug Administration will have to change their name to something else.
What about the other way around? Can someone who works in the private data sector leave and go to work for the agency?