PandaBot: A friendly, affordable 3D printer | Hacker News Reader

PandaBot: A friendly, affordable 3D printer

kickstarter.com

51 pointsnerdburn44 comments

My good friend Kelly is finally rolling out his inexpensive ($800) 3D printer for everyone. Please support the Kickstarter.

Open articleView on HN

The video mentions they've only been working in the 3D printer world for less than a year. So now they know how to fix "all the problems" to make 3D printing available to the masses? The leadership page doesn't give me a lot of confidence in that blanket statement.

Regular 2D printing isn't something that's been easy for most people after 20 years of modern computing. It's a toss up whether or not software, hardware, or user error.. For that reason, I think 3D printing will be a hobbyist thing for another 5 years at least. The delta between a PandaBot and a high-quality MakerBot 2 with a good reputation is only about $1,000. Thats not much for a die-hard hobbyist, or rich geek wanting a new tech toy to fiddle with.

The MakerBot 2 is on its second iteration, designed and built by a thought leader, and looks really fucking impressive compared with the 2-tower and unprotected print head design of the PandaBot.

When people want to bring devices to the masses, they need to consider what that means. How many kids may stick a pencil or finger into the moving parts when Dad isn't paying attention? Or what about my maid who's spraying dust repellant around my desk? How are you going to fight warranty issues on things nearly impossible to prove, and for which you're not protecting against via thoughtful product design?

In effect, you've reduced costs to make the PandaBot accessible to a class of people who aren't really your customer, and that cost-cutting meant the removal of key parts that truly make a device ready for the masses. Like protecting the 'auto calibrating' head.

I think it's great you're working on that goal but expectations seems a little disconnected from reality (from my perspective).

In effect, you've reduced costs to make the PandaBot accessible to a class of people who aren't really your customer

And you know this how? I've been surprised at the consistent hostility of some people on HN to innovating on price. This is the same mentality that led to the US car market going temporarily insane and assuming that everyone wants the biggest SUV possible.

It's because, in practice, most so-called price innovation is of the form "I just bought 2 million sticks of butter for $1 million. I think I'll sell them for 25¢ apiece." Being unprofitable or even just barely scraping by is not actually innovative, but too often that's what passes these days. (I'm not saying PandaBot is that way. I'm just offering a reason for the pushback against "innovating on price.")

That wasn't fair. Pricing so close to COGS that you can't sustainably deliver the product isn't price innovation, nor is pricing so low that COGS forces you to deliver an inferior product; the comment you're responding to took pains to spell out why this particular 3d printer price didn't make much sense.

The consistent hostility towards low prices on HN is based on this site's manifest tendency to underprice things, or to calculate price from cost or from pie-eyed notions of how big their addressable markets are.

Product Manager again.

Total units shipped is incredibly low in 3D printing such that there are very few economies of scale. A certain Brooklyn firm has shipped 13,000. Ever. They have massive share of total and annual global shipments, including the highest of high end industrial printers.

We think that the technology is ready to drive this to a seriously large audience. Our COGS benefits from using the global supply chain and designing for manufacturing. It also benefits from scale. Sperry-Burroughs used to hand assemble a few products, then the world changed.

We're not priced to hand assemble in Brooklyn and would never make money doing so. But there are other, very high quality and proven, options.

But Tom, what basis do you have for suggesting they can't sustainably deliver the product? I don't know whether this is the case, I'm waiting to find out. But I'm perplexed by the assumption that they can't, and thought the grandparent comment assumed its conclusions. I don't agree on the inferior product angle; that assumes someone else should define what the minimal acceptable functionality is.

I don't necessarily think the PandaBot is directly competitive with the Makerbot- I'm sure corners have been cut to achieve that low price. But I don't necessarily need everything that the Makerbot does. It seems to me that as a technology is refined, you have the choice of adding more features, or keeping more limited features and lowering the price - that's what I'm referring to as price innovation.

Consider that the Thing-o-matic was $1200 or so, and people were happy with that at the time of release (3q 2010, IIRC). So the new Markerbot is a lot better, but it's also nearly twice the price. The PandaBot is offering something similar but simpler for $800-ish, which doesn't seem wildly unrealistic to me given the 2-year interval. I can see from the Kickstarter page that it's more of a no-frills product - one which requires more user supervision and safety-awareness, for example - but the tradeoff is that it costs less. I'm not sure whether I'd buy one (since it would be more of a hobby than a tool purchase), but being under $1000 is a big factor in any purchase decision.

Yeah, I don't know. You make a good point too. Maybe the parent comment was worded too stridently, but I thought it asked a valid question, and I think HN needs that question to be asked more, not less.

There are a number of Reprap variants that print better than the Replicator 2 for less than half the price.

Examples?

The Prusa Mendel sells for $799 on Makergear and can print at 1/10 the layer height of a Replicator 2. Thats the same price as the Pandabot (well, give or take $1), the advantage of the Pandabot is that it auto-calibrates and is excessively easy to assemble.

Although the official spec lists a 100 micron layer height, technically, I see no reason why a Pandabot couldn't print at the same resolution as a Prusa Mendel. Having seen both a Pandabot and a Prusa Mendel in person the electronics and hot end are similar enough that there is really no reason they couldn't achieve the same 10 micron resolutions on the PandaBot.

Sources: - http://www.makergear.com/products/3d-printers - http://blog.reprap.org/2011/12/001-layer-height-on-prusa-men...

>The video mentions they've only been working in the 3D printer world for less than a year.

The design lead said that he's being working with 3D printers since 2008. It's at 30 seconds into the first video.

Exactly, this comment should really be higher up.

Liav, the designer of the PandaBot, was one of the first people in Canada to build a RepRap Darwin. The year Panda Robotics took to develop the Pandabot is the result of many years of prior 3D printer experience on Liav's part.

> Regular 2D printing isn't something that's been easy for most people after 20 years of modern computing. It's a toss up whether or not software, hardware, or user error.. For that reason, I think 3D printing will be a hobbyist thing for another 5 years at least. The delta between a PandaBot and a high-quality MakerBot 2 with a good reputation is only about $1,000. Thats not much for a die-hard hobbyist, or rich geek wanting a new tech toy to fiddle with.

This a very interesting statement. Couple it with the news that HP will not be profitable till 2016, [1] why doesn't HP slap on an extra axis to their printers and make them extrude ABS instead of ink? Granted, they might not get the demand they want in the short term, but they are setting themselves up for something bigger later down the line. They have the infrastructure, distribution, know-how, and cash. What's stopping them from cornering the desktop 3D printing market?

[1] http://arstechnica.com/business/2012/10/hp-ceo-were-screwed-...

HP actually has IP around 3DP and until recently sold Stratasys machines under their private label. I say recently because it was shown that HP sucked at selling them and ceded its territory back to Stratasys.

Also - some companies (ZCorp / ProMetal) use HP print heads to deposit less viscous materials for powder bed 3D printers

So I went to the Stratasys site and saw that there entry level printer was $10k, would this be the equivalent of a Ferrari vs. a kit car conversion?

Why doesn't Stratasys make a Honda Civic of 3D printers?

From their perspective, $10K probably is the Honda Civic of 3D printers. They probably see all the low-cost stuff as "toys" that aren't usable for "real work". It's just the innovator's dilemma part infinity.

To Stratasys $10k is probably already the Honda Civic of 3D printers.

To go with an entirely different analogy, Stratasys is the old, lumbering baron of 3d printing. Conversely, RepRap, Makerbot, Ultimaker, MakerGear, Type A and now Panda Robotics are all the agile, new kids who were raised on a healthy diet of open source. Having invaded the Barons lands they are now zipping around rearranging the Baron's fiefdom at 100km/h and generally trying every way of doing everything all at once. My bet is firmly on the new guys, how they'll divide the spoils once Stratasys has been finally overwhelmed by all this chaos is the bit I'm curious about.

This comment will probably get some flack, but it sounds like they need to get bought out by a PE firm and go through a massive reorganization. Layoff all the top management brass, keep engineers that have a low-cost printer in their homes and hire someone like Bre to keep things lean.

Otherwise, they will watch their market share erode until one of the scrappy startup kids get enough bank to buy them out for pennies on the dollar.

Or more likely, a PE firm would saddle the company with debt to pay off the new owners' debts, lay off everybody with any know-how, make an inept attempt at reinventing the now low-morale company, then give up after a few years and go into bankruptcy selling it off for parts leaving a pile of bondholder and pension lawsuits behind while management moves on to their next glorious MBA adventure.

Yea, it's a good question about what Statasys' response will be, but in the near term they see, realistically, that much of their customer base never heard of the other FDM printers. They've got a sales force and install base, so I bet they're good for years to come. Still tho - they're mostly likely to be disrupted in the space.

Honestly, the editor of Modern Machine Shop hadn't heard of Make Magazine when I asked - the maker hobbyists and manufacturers have little overlap surprisingly.

Architecture student here.

There is actually a large consumer base for a product exactly like this. I know many other architecture students that have spent more than the cost of a PandaBot (as well as months of hard work) building a RepRap printer to keep on their desk and use in studio. And we keep things (like models) far more fragile than this printer on our desks all the time.

Halfing the price of a comparable printer (MakerBot) is in fact a huge price difference (and therefore, a huge motivator for purchase) for people like students.

Only $1,000 difference? Makerbot's Replicator 2 is approaching triple this price at $2200. Makerbot is moving upmarket into the semi-proffesional market and abandoning the hobbyist and student market leaving an opening for sub-thousand dollar robots like this one.

In fairness to Replicator 2, the output versatility looks scores more impressive than Panda's: http://www.youtube.com/watch?v=3o6pcbhylmQ. I suppose it all depends on the hobbyist.

The Ultra-Bot 3D Printer Kickstarter ends in less than 22 hours and has a full assembled unit at $899 ($99 more).

It's based on the original Makerbot Cupcake but upgraded and improved from trial and error. And there is a version with a larger build area (8x8x8"), which is the one I've pledged to receive. And with the heated build platform you can run ABS or PLA.

The thing that bothers me about the PandaBot is the lack of details on the build area size (update: 11x11x11" Print Envelope" and heated platform). Also the price isn't that great compared to other fully assembled 3D printers with the same tech. Also I'm not a fan of the base moving instead of just the print head. I can't think of any wide format printers that move the base instead of the entire head assembly/gimble.

I do love the snap together setup process and the look of the machine. It looks like something you would buy in a store and almost anyone could setup and start using immediately.

The build area is 11x11x11"

The snap together process and self-calibration make it a much more useful experience. We've been bringing it all around town and usually have it set up and printing in minutes.

You can especially see it in our BNN interview. Liav had to set it up in minutes because he couldn't be on set any longer than that. :)

Full disclosure, I'm Kelly John Rose of Panda Robotics.

Thanks for your comments.

We've found that moving the bed horizontally helps reduce the size of motors necessary to move the head as well as simplifies the required inertial calculations to properly model the fluid flow to give a detailed, low defect object.

We're definitely aiming for a store bought look. We've violated the sacred law of demos time and again and not yet been caught. To do Canada's CNBC we dragged it out of a cab and had it set up and printing in 5 minutes with no adjustment, with an alpha prototype! On live, national TV! So it's a decently robust design.

awesome job on the look and ease of use and good luck on the project.

I love 3D printers, but this is the third one I've seen this week in addition to MakerBot and Form 1. I'm really not understanding what the differentiator is after seeing the video except a slight change in price.

This is a FDM printer that is a lot cheaper than the MakerBot ($800 vs $2100). The Form 1 is a SLA printer, which means it is better at making small, precise parts; but worse at making parts that are durable enough for actual testing in assemblies (as opposed to just looking pretty).

It is interesting to see competition to the MakerBot, since $2k is still a bit high, and if the PandaBot's "calibration-free" design works as well as they claim, the ease-of-use factor is still a big thing to exploit.

If it has even moderately good print quality, a commercial, fully-assembled FDM printer for under $1000 is an interesting prospect.

Thanks.

The basic thought was to get it to a point that we could provide it to designers and university students with little to no engineering experience and they could start printing their 3d models right away. Hence, the heavy emphasis on self-calibration and ease-of-use.

The price range just works because we also designed it from the start to be a manufactured product, so we get an economy of scale very quickly as more people pick it up.

You can see print quality examples on ponoko, Derek swung by our office to see it close up in action yesterday: http://blog.ponoko.com/2012/10/04/pandabot-3d-printer-launch...

We think price is important and that the true power of 3D printing will come when we break out of the niche and into the mass market. Beyond price, we're designed from the start as a manufactured product, which led to much different component selection and engineering choices. These different design decisions gave us a much more rigid, robust, and quiet machine, while also allowing us to leverage the economies of scale and manufacturing skill of the global supply chain.

I'm running product at Panda Robotics.

Speaking of rigid - what's it with the heatbed (?) bending to the left as soon as it changes direction in the video (at around 0:50 - 0:55)

It is prototype hardware and some elements are not where we'd wish them to be. We have a fully revamped design for the bed carriage and drive assembly.

Even though it's 3x the price, I'd take the Form 1: http://www.kickstarter.com/projects/formlabs/form-1-an-affor...

3x the price and 10x the machine.

I'm waiting for the 3d printer you download and assemble entirely from 3d-printed parts.

And it will be made out of lego, or at least lego-compatible construction system.

I'm waiting when you can get plastic in a form that works with these things at $2/pound.

There has been a lot of work lately at creating filament makers that would allow people to recycle scrap plastic into filament you can print with at home. I think that when this technology starts to mature things ought to get rather interesting.

This is what the rep-rap project is working towards.

One of the first obvious issues is how do you print motors? How to print circuit boards (with 3d printing hardware that is) is another obvious question, though that seems a bit more approachable. Wires are also another obvious one, though printing internal leads into pieces seems like a decent way of avoiding wires after you have gotten the "print a circuit board" thing down.

Less technical issues are "Why do you even want to print the entire thing?". Right now most rep-raps use many metal rods. These could be replaced with printed parts, but is there really a reason? The metal rods are cheap, easy to source, and reliable.

This all said, what rep-rap has accomplished to date is pretty impressive; they have snatched up most of the low hanging fruit and most of what is left isn't stuff that makes sense to print anyway.

Anyone else remember having to hand-calibrate 2d printers?

Toronto based company!

Yep, just across from OCAD and AGO. :)

  1. What is your firmware based upon?
  2. What computer software do you use to assist PandaBot? Slic3r?
  3. Do you need a computer hooked up via USB to facilitate printing?
  4. Do you publish schematics and/or physical part dimensions for repair 
  and replacement?
  5. What makes your product better than a RepStrap?
  6. A MendelMax full kit costs $699 USD. The kit is complete with full build 
  instructions and fully repairable if/when something goes wrong. What makes 
  your printer better than this model?
  7. How do you handle inertial jerk when you have a partially printed model due 
  to the moving table?

1&2 We are using available firmware and software with the prototype while we build our own versions ( fully respecting the licenses the community is using for different projects). We're building software and firmware that are properly paramaterized for our system and give great results - you need to get the exact fluid mechanics right for your hardware. But it will be totally hackable and you are more than welcome to roll your own g-code with your favorite tool or even pick your firmware.

3 We load files via USB but you don't need to keep the computer plugged in after a successful transfer

4 We're generally using off the shelf parts but it's a commercial shelf. Trade-off of quality, tolerances, capabilities. We feel that the improvements to the stability and reliability of our product are worth going in this direction.

5 We're building a tool to let people focus on designing and producing physical objects. Different design objective than kits. By not being a Rep* we get access to interesting components with very tight tolerances which will give users a wonderful experience in reliably producing objects.

6 We're shipping a manufactured product, it comes in 2 pieces to reduce shipping costs. It's a tool that plugs in and goes. Auto-calibration and the incredible rigidity really help make it "just work".

Again, it's a different focus from kits so different strengths and weaknesses. If you want to build your printer from scratch, we're not the right choice. If you want to be printing out of the box with no calibration, re-calibration, or learning all of the details of Slic3r or other tools, then we are the right choice. And then, if you want to explore Slic3r or optimal 3D material strategies without having to become a hardware/robotics guru, we are the right choice.

Lots of room for different people to focus on different things, or for people to have different tools for different projects.

Reply on news.ycombinator.com