Anytime have experience with starting a printer farm? What would be the advantages of going with something like this over diy?
http://www.typeamachines.com/series-1-pro?__hssc=&__hstc=&__hsfp=2103173164&hsCtaTracking=a1cbcd01-7d82-4455-a81f-4618bc95b8f5|e80e61fb-09a1-462c-b9ee-c60a091b88b7
I’m assuming by printer farm, you’re talking about setting up a bunch of printers for manufacturing purposes?
If so, then yes, I have experience with that for sure. (Possible disclaimer, I work with several groups that own print farms). The major advantage is the ability to quickly change the design, and when you’re not producing that many parts the time cost makes sense. Depending on how your business works, you might have quite a few printers running before you decide a part is worth injection molding.
@Justin_Nesselrotte thanks for responding. My question was about this product verses building printers myself. The price of these machines seem way more than it would cost to build. Anyhow what do you mean that you have experience with printer farms. Is there a group I should know about?
@Ryan_Besch for someone that knows how to build a printer, is it worthwhile to spend this kind of money for their support.
@Dovid_Teitelbaum there’s not really a group, but I’m the developer for BotQueue which is what the clusters that I know about run.
@Justin_Nesselrotte I just heard about BotQueue. I understand it’s open-source. Do any of the printer farms work for outside jobs or mostly in house stuff?
@Dovid_Teitelbaum most of them are in house, but a few are getting set up to start with outside jobs. They would still work through things like 3d hubs though.
@Justin_Nesselrotte I was told 3d hubs isn’t so great as your working with individuals that aren’t really interested in paying for stuff as apposed to businesses that realize the benefits of prints and have an ongoing demand for stuff.
For a print farm you want dependable machines. They need to operate with the minimum of adjustment and perform reliably.
While you can self-build machines you might be tempted to use cheaper or lower quality components than are used in pre-built machines.
Regardless of support from the printer manufacturer you pay a premium for quality and reliability in a machine designed for heavy use.
For a print farm which is intended to make money it makes sense to invest in reliable machines. Downtime will hit your profits and reputation.
I have an 6 bot farm, (it’s been up to 8) 4 for production PLA parts, 1 for hi temp and exotic material and 1 as a testing mule, extra capacity, prototypes. I’ve been printing commercial parts for a few years now. Two of those production machines typically run 6 days a week, 8-15 hours a day depending on demand.
If you have the time and ability it’s possible to build and configure your own farm. Building, configuring and deploying them is a skill in and of itself. As is being able to get the slicing right and get quality prints. Your time to market can be shorter by buying a complete machine. I can also scale back down by selling the machines when demand drops by cycling out the older machines or parting them out.
If you have the funding and the lead times are minimal for for the machines I’d buy them. One advantage I have is I sell kits of what I use in the farm (with the exception of the hi temp machine, I built that a couple years ago) so I can scale up in a few hours by building in house. If I were just going to be a parts house, I’d buy a fleet and be done with it. My starting choices would be TypeA, Lulzbot or Ultimaker with the final decision on what the machines were going to be doing in terms of material type, number of extruders, build envelope, etc
Since Printrbot printers are less expensive, they make decent bot farms. We are building an example farm to show how cheap and easy it can be. The intention is to demonstrate how densely packed our machines can be to use minimal space. We will use the farm to build our Botfarm software we have on the roadmap. We will probably end up with two options- run it through our cloud for many members to print to it- auto queuing the models on the available printers automatically… Slicing in the cloud, and a local server version to run at home if you want. The cloud version will also work for a single user if you prefer.
The last goal is to build a robot that uses removable build plates to remove the prints, rack them, then load an empty plate. If we succeed, the farm could have very little down time.
A man can dream!
Brook
Printrbot
That is an Awesome printer if you can afford it?
Yes, I love the Type-A guys! It prints beautifully.
Type-A Machines seriously caters to commercial printer farms, it’s their core target market. It’s a great choice if you can afford it. (And if you think you can’t afford the TAM, I bet you REALLY can’t afford the troubleshooting and downtime you’ll get from building your own for cheap.)
@Brook_Drumm Watch out for Makerbot’s automated build plate patents, they’re kind of idiotically broad. Basically any way you can imagine moving the build plate in/out of the printer will infringe. The unattended-printing build plate solutions I’ve seen that don’t infringe all pop off the print from the bed without removing the plate.
I haven’t considered patent infringement. Bre asked me one time “what’s the end goal?” and I said " to beat you! 
… And not get sued." He said " we’re not gonna sue you." A funny exchange. He was actually really a nice guy in person, btw.
Since the robot is separate and not an integrated part, and any robot arm could remove them, I don’t think it infringes anyway. Two separate things. A printer (any printer). Any removable print plate (widely available across many companies).
Brook
@Brook_Drumm We may see Stratasys attempting to use the patent hammer more and more as their market share is eroded by consumer-grade machines. I hope not but time will tell.
The problem you’ll run into with Makerbot’s ABP patents is that there is some very broad “method patent” language. Check claim 17 of US20120059504A1 - Automated 3d build processes - Google Patents for what I mean. Basically this patents any means of executing an automated sequence that does the following:
- Print an object
- Cool the build plate
- Remove the build plate from the printer
And claim 18 covers doing #1-3 plus: - Put a new build plate in the printer
- Heat the build plate
- Print a second object
The way the method claims are worded are very broad. You could probably fight it in court by saying that manual swapping of build plates is prior art. (Method patents can cover manual operations too.) But do you really want to risk the legal costs to fight it?
what about a conveyor belt?
@Dovid_Teitelbaum A conveyor belt is actually what MBI was intending to patent. I count three different patents related to conveyor-based build areas, their operation, and heating/cooling.
Related sidenote, Charles Pax originally invented the Automated Build Platform (using a loop of Kapton in a conveyor arrangement), so Makerbot hired him and immediately patented his invention. Which is legal, but sketchy. In my opinion the ABP patent was one of the few truly crummy things Makerbot has done with end-user IP, since Charles wanted the design to be open source but his contract allowed MBI to patent his previous work. (Whereas most of the other “Takerbot” stuff is seriously blown out of proportion.)
I still don’t think it infringes since the robot doesn’t need to issue any commands to the robot. The robot just watches and performs the operation. They didn’t patent a removable build plate, or a removable print, or the fact that something must remove the print… Just a printer that does it all. Two autonomous systems working along side each other isn’t the same as one system doing two things in succession. I’m no lawyer, of course. But I gotta believe that a robot arm that reaches in and removes a plate on its own is just a robot. Not a 3D printer. It could also serve you earl grey tea.
@Brook_Drumm that’s why it’s a bad patent. (As in, too broad, probably wouldn’t hold up in court.) It DOES appear to claim any use of a removable heated build plate. There’s prior art for that. Just make sure you understand your worst-case exposure.