June 23, 2014

3D Printing - Step Four - The Future

Before I launch into an argument with myself and a larger community, first an apology on the length of this article, get your cup of cocoa ready. Right, Lets prepare an analogy - The printing of the written word.

In the beginning, possibly just after the earth cooled and the dinosaurs came, man daubed walls with mud and stuff. A very crude story telling tactic but some still survives today. Once words and ideas were formed, we invented tools and moved to stone and chisel, ink and paper etc. Mass printing came along with the printing press but it's use was limited to a few. The break through for individuals to write lots came with the manual and then electric typewriter... well, you could argue the secretary came first, but that's just nit picking.

It wasn't really until the teletype that an individual could actually send a "document" to be printed in a "mass" fashion, with wonderful type quality to boot if you didn't mind the green and white striped paper. Dot matrix printers were next with rough output but colour to the home user was a real win - anyone in my generation will remember the Star LC24-200... well, hopefully. Printing then took a bit of a split, laser for business and inkjet for the home, and this is where we have remained for the most part. Sure colour lasers are available for the home, but the are expensive, and you can always just nip into the office to print a colour booklet out... erm... not that I've ever done that of course.

The Download Generation

What's this got to do with anything I hear you mutter, well, the path of getting a thing in my head to an object you and your mate can own has taken much of a similar path. I could make it out of Mechano or Lego, or whittle it out of a stick, I could make it out of clay and give it to a manufacturing company, or, now, I can prototype it myself before sending it off for further refinement at significant cost reduction, or I can just send you the prototype. There is a huge difference though, the tools we developed as a society to disseminate the written word were relatively easy to use so the masses could use a keyboard quite easy. Learning to use Blender - a 3D modelling package - is kind of like learning to unicycle across the grand canyon on a tightrope while juggling fiery chainsaws. In fact all the 3D modelling packages out there are complicated by average ability standards, and here is the first barrier to adoption as a generic, consumer technology.

Sure you could get your mate to help you, or a design house, but really getting the exact thing you want will be tricky and getting someone else in will generally cost you, hopefully just a couple of Bourbon Creams but usually significantly more. You can download what other people have done and in some cases tweak to your needs, and this is really the leveler here. The area of mass production at a local scale, instead of mass producing, say, cutlery in a factory and shipping it to distributors to store and sell, you could download the files and print as you use them. Innovation however, would still remain the niche of the fiery chainsaw juggling few.

Lets stick with the download market for a while. Chainsaw manufacturers produce chainsaws with sacrificial parts - so that if the chain jams you don't shred the motor for example. Instead of waiting a few weeks for Black and Decker to ship you a part from their warehouse across the planet, they could email you the file and you print it out locally in a couple of hours.

Oh, I think it's probably a good point to mention I don't really have a chainsaw fascination... well, maybe just a little one, but I had a huge garden that required some clearing... Anyhow, I digress.

It's at this point I should also introduce another analogy to demonstrate "tipping point". Autogas. A bloody good idea, but cars had to be converted and only a few garages provided it... Car owners won't convert until there are more garages available, and garages won't produce more stations until there is demand to cover the investment. Both sides of the issue need to approach the middle at about the same point. The same is true here, Black and Decker will not change it's business model until enough people want to/are able to use this revolutionary way of distribution, but to use it, people need a decent printer so we are back to the uptake of the technology. In order to stimulate this we need to put the technology in every house, as well as produce more things in biodegradable, non-petroleum based materials.

The Technology

Back to my written word analogy, it was the birth of the cheap inkjet that put a printer in every home with a computer in. Where are we with 3D printers? well I think we are just getting to the end of the dot-matrix revolution. The industry is currently split, Fused Deposition Manufacturing (FDM, FFF, FFD) for the home and Stereolithography (SLA, SLS) for the business. The biggest problem at the moment, as it was with early regular printers, is cost. Any printer you want currently is not a cheap thing unless you get a kit and spend a week putting it together, who has the time and patience for that. A decent FDM machine will set you back at least £1,500, and SLA machines with a tiny print area are about £4,000 with industrial units costing 10's of thousands of pounds to buy and run.

Sticking with the home use market for now, the next issue is that these printers are VERY slow. Printing something the size of a fork takes over an hour and if you want a decent resolution you can easily be waiting 2-4 hours. Printing a large mug ranges from 12 to 114 hours. Speed will increase as the technology matures, just as we went from 6 minutes per page with dot matrix to 6 pages per minute with inkjet. If you combine this with quality, you end up with the loosing end of an unpleasant stick, but if the speed is good enough you can hopefully use some of that speed to improve the quality of your final print.

Clunk and noise probably come next, but designs will improve though so this shouldn't be too much of a worry. Apple will eventually invent one and all our worries will be over... erm...

The biggest hurdle that general product evolution will not overcome in itself is that of control complexity. As I highlighted in my previous post, there are so many parameters to fiddle with, and depending on the actual material you're using settings can also change dependent on colour within the same material. In a similar way to inkjet printer manufacturers producing 'certified' inkjet cartridges for their printers, it's not beyond the realms of possibility that a printer will identify the material it has loaded and then analyse the model it's sent to work out the best settings. How will this work? You can currently buy cheap, generic PLA that works but it's not that great. You can buy higher quality 'branded' plastics from ColorFabb, Faberdashery, FormFutura etc, but they are a little more expensive... to get printer brand certified filament, won't prices rise? Will the fact that everything is open source keep this in check? Will manufactures start to produce a 'standard' filament? There are so many moving but connected parts here that this will take a while to sort itself out.

If your printer is hooked up to your broad band though, not only could you control it remotely, but it could go to the material manufacturers site, identify itself as a printer and download the settings for that material and printer combo. Revolution my friends.

How far off are we?

A leap into the world of Arthur C Clarke and Tomorrows World, and their predictions of the year 2000. Well, I can see a fully automated SLS based machine that uses powder, meaning it's self supporting so any model can be printed, and an automated clean up and recycle of unused material. I can also see how you could extend the materials to things like chocolate which would be a wonderful 'gimick' to get the technology out there. If you could cool the chamber and speed the laser bit up, you could get though a layer quite quickly, so speed would be better as well.

The problem with SLS technology though is that you need enough powdered material to fill the print bed. Is this the way forward? Enough chocolate powder to cover an A3 bit of paper a foot thick... I can feel the diabetic coma approaching already, if one were to accidentally spill the container while having a munchie attack.

Star Trek Replicators then?

Why not. Clarke and tomorrows world predicted fully automated, humanoid helpers and hovercars by the year 2000, so replicators are equally likely by the 24th century.

The next 5 years then?

Well, lets look to the past. RepRap was the first usable hobbyist 3D printer in 2005. It took 3 years before someone thought to print an actual thing that might be useful on it. 9 years on and the technology has refined, but more slow evolution that rapid revolution. While there are large marketing campaigns by people like Maplin, these only really appeal to the techy folk. Companies like Creat3d, iMakr and Help Anyone Create are popping up in the UK to help consumers, but uptake is going to be slow due to the limited use, expense and complicated software.

I think this technology will remain niche hobby for a while yet, we will have multicoloured print heads in the next 5 years, we already have temperature reactive plastics and printers with multiple heads so extending these out will be pretty simple. while the technology remains a whizzing print head driven by motors and fed from a filament through a heated nozzle, speed will be slow to increase. Resolution in the X/Y axis will increase, it's already possible to use a 0.2mm nozzle with some printers, but speed is the key, You are also more likely to get a jam with smaller nozzles, especially with the more exotic materials like the wood fill.

Do you have any optimistic outlooks?

Why yes, thanks for asking. Not for the average hobbyist for a while though. Business will drive the technology for a while yet as it did with laser printers. Visualising things is a key part to a lot of businesses, floor plans, gadgets, gizmos, cars... all require some prototyping. Being able to tweak a thing and have it print overnight ready for review in the morning is a definite possibility. The use of inkjet technologies spraying heatproof ink into the material before the laser sinters it is already out there. Improving the speed and quality will also happen.

It took 40 years to develop the transistor, and another 20 to minaturise it to a point where it's useful in anything finger sized so the amount of time it took to develop an iPad was actually quite small in itself but it was developed on the shoulders of the transistor. The philanthropy involved in Open Source and 3D printing will make up the time to develop the transistor (and other components) into an iPad but I reckon it'll be a good 10 years before 3D printing is a mainstream hobby.

Overnight printing is already here, these all took around 10 - 14 hours to print:

I think the most likely 'practical' use of this technology will be reconstructive surgery. Most humans have spare bits so as MRI technology and resolution improves, this data can be used to recreate bone structures in case of loss or disease. Humans are not symmetrical but close bone structures to those you had before the loss will improve the time it takes to relearn to use the affected part. In small scale, SLA technology is already in use for dental reconstruction.

Another area that will benefit is archaeology. Scans of various famous mummys were done several years ago, now these can be recreated to see what they really look like. Any type of depth information can be used, so ground penetrating radar for example. Even sound reflection data to model continental composition can be used. You would obviously have a bit of a problem getting the nearest continent on your print bed though - best not to try that at home.

Coming up in the next episode...

The next post will predominantly be about 3D scanning. I will be working with a few people over the next few weeks to improve my scanning ability and then there will be more heads like the above and hopefully some desktop art, plus lots more photos.

UPDATE: and here it is.