August 19, 2016
Andrew Boggeri of FSL3D discusses the future of 3-D printers and predicts the rise of marketplaces bringing together designers and buyers.
Andrew Boggeri is a lead engineer at FSL3D, which manufactures high-quality consumer desktop 3-D printers featuring laser-based stereolithography technology.
PwC: Andrew, can you please tell us about your company and how you got into 3-D printers?
Andrew Boggeri: Sure. FSL3D is a division of Full Spectrum Laser, which began in our founder’s garage in 2010. Full Spectrum Laser started out reselling laser cutters from China. Our founder realized that although the mechanics of those systems were OK, the electronics were not well engineered and the software was almost impossible to use. So we developed our own high-power embedded circuit board to control the lasers, and we developed our own software front end. We also sell customdesigned, high-end metal marking lasers.
In doing so, we gained valuable experience in creating a control system for galvanometers, also called galvos. Galvos are devices for steering laser beams. Each one is essentially a DC motor with built-in feedback that has a very light shaft attached to a very small mirror. The system has minimal rotor inertia and so can be used to steer a beam at very high speed with excellent accuracy. A tiny deflection of the mirror will produce a very fast movement over a great distance.
We saw that 3-D printers were really popular, but all the desktop ones were using FFF [fused filament fabrication] technology. We’ve owned several different types of 3-D printers and we have used 3-D printing often during prototyping. Most of our experiences were neutral at best. Given our deep experience with lasers, we saw an opportunity to create a laser-based desktop 3-D printer. We launched a Kickstarter campaign to raise $100,000. We raised more than $819,000 in 33 days. Pegasus Touch is our desktop stereolithography [SLA] printer.
PwC: What made it possible to create a laser-based desktop 3-D printer at the price range you are offering, which is around $2,000?
Andrew Boggeri: A big portion of the cost for laser 3-D printers is the laser modules. The tremendous change recently has been the proliferation of Blu-ray players, which brought down the cost of laser modules. Blu-ray diodes are now available at a lower cost, making it possible to engineer a desktop system at the price points we are targeting.
We originally looked at DLP [Digital Light Processing] projection technology; the technical term is MIP stereolithography, or mask image projection. With MIP, instead of moving a beam around, a full image of the layer is projected so the whole layer can be cured at once. We found this approach very slow, however. Because the technology uses a light bulb and the light is greatly diffused, it takes several seconds per layer to cure the resin. We had built a prototype, but we shelved the project in favor of using lasers.
PwC: How does the performance of your printer compare with the high-end printers?
Andrew Boggeri: First, price is a big difference. Where we excel against the high-end printers is in the speed and the ease of use, and we are comparable and often better in terms of quality. The high-end printers use a different class of resins that have entirely different chemistry, and require a lot more expensive laser diodes to cure the resin. The tradeoff is that you can build totally transparent lenses with the highend printers. You will also get high fidelity and some better properties out of those resins. We searched very hard for improved optics and now have a system that is comparable with products that cost more than 10 times as much.
PwC: Most printers in use today work with only a single material. What would it take to make them multi-material printers?
Andrew Boggeri: This is a difficult problem for 3-D printing. Our particular technology, SLA, is a single-material technology. Conceivably, it could be turned into a multi-material technology by adding more types of resin and lasers. Such additions would make the printer very complex, and there would be a lot of operational tradeoffs.
FDM printers will be more advantageous for multi-material printing, because more heads can be added at much lower cost and complexity. However, the multimaterial printers getting the most play are using Voxeljet technology. It’s similar to an inkjet printing head with drops of material. For multi-material printing, I think that is the present and the future.
PwC: When will we see desktop printers that print with metal? We have heard of the ones that use arc welding to print with metal.
Andrew Boggeri: I don’t think we will see a desktop-size metal printer that uses the direct metal laser sintering [DMLS] method. These printers use fine metal powder that is bonded with a laser. Two big barriers are stopping DMLS metal printing from coming down in price. First, the lasers needed to melt metal are typically highpowered fiber lasers. Today the least expensive of these lasers cost $15,000 to $20,000 for an entry-level model, and the entry-level models are not powerful enough for DMLS. Second, the cost of the raw material—the fine metal powder that the printers need—runs anywhere from $600 per kilogram and up. However, DMLS can produce very small features, and therefore complex geometries are possible.
For now, the downside of a weldingbased desktop printer is that the bead of a weld is on the order of millimeters rather than a few microns. So it’s suitable for products that don’t need high resolution.
PwC: One barrier often cited is the slow speed of 3-D printers. A print cycle can take hours. What are the prospects for printers becoming faster in the future? Are we anywhere near the physical limits of what is possible?
Andrew Boggeri: There are lots of ways to improve speed by using higher-quality components and optimizing the designs and movement of the lasers. Of the laser desktop printers, we’re among the fastest right now. With our control electronics, we can pulse the laser at 500 Hz while printing at 3,000 millimeters per second.
The slow part of the SLA process is the release and recoating cycle. One must either withdraw or tilt the head to get the resin to release and then recoat the next thin layer of resin that will be cured. We’re researching some special release coatings that the resin adheres to much less than our current coatings, so we can perhaps double—maybe triple—the speed of this cycle and still maintain the quality.
The other factor that impacts speed is how fast the resin is cured. We could drive our lasers faster and cure this resin at a very, very fast speed—probably 10 or 20 meters per second, rather than 3 meters per second like we’re doing now. The chemical reaction proceeds much faster than we’re able to drive our lasers.
In the filament printers [FDM], speed is influenced by the particulars of plastic melt rheology. Today, a fast printer is moving about 300 millimeters per second while it’s pushing plastic out. If you try to move faster, you will end up with a bunch of Hershey’s Kisses on the ground. So there are likely physical limits to how fast you can move the head.
The technology that has good headroom for speed improvements is inkjet-based printers. You can deploy a lot of heads to work in parallel on a given layer. It’s really a matter of cost and complexity that rises as the number of heads increases.
“There are lots of ways to improve speed by using higher-quality components and optimizing the designs and movement of the lasers.”
PwC: Are you seeing any challenges to adoption that the industry needs to address?
Andrew Boggeri: The 3-D printing ecosystem depends on digital files that represent the designs that would be printed. The problem that still exists is how do people buy or sell these designs? Today consumers can buy an STL [STereoLithography file format] file from several online libraries, and it is not uncommon for designs to cost $40 or $50 or more. Highend sculptures cost even more, perhaps $600 for a print or $2,000 for the design. Today the designers do not have any protections, as there is no built-in way to protect an STL file from being copied and reused. This lack of protection is reflected in the high prices of designs. That is not the future.
What is needed is a marketplace like Apple iTunes, where 99-cent impulse purchases are available. At these price points, customers can try designs. That’s what we’re doing, and soon I think everybody will. We have the touch screen on the printer, so the design library store will run on a printer without the need for a PC. This store is going to work like a combination of Amazon and iTunes.
A marketplace will connect design creators and design buyers. Our printers will have built-in digital rights management [DRM], and every printer is encrypted to an individual serial number. This software model will be available like an app, in that a designer can sell a copy of this model to anyone who has this printer and the model works only on that individual’s printer.
Without the threat of designs being stolen and copied indefinitely between printers, design owners can price them much more affordably, such as 99 cents or $5. Lower prices should increase adoption by more designers participating in the ecosystem.
“On a longer time scale, beyond five years, the bigger trend will be the rise of midsize companies that do not need to own their own manufacturing base.”
PwC: What do you expect to see in the future that is not here today?
Andrew Boggeri: I expect the rise of distributed peerto- peer buying and selling. There are digital crafts and digital artisans—people who are creating digital things using modern technology. The 3-D printers are the manufacturing base for these things and will enable folk manufacturing. In small ways it is already happening, and it will accelerate as the price drops.
On a longer time scale, beyond five years, the bigger trend will be the rise of midsize companies that do not need to own their own manufacturing base. As 3-D printing tools become better and more automated, I can imagine a day where people buy just a circuit board for their new phone and print the rest at home or at the nearest 3-D printing center, customized to their needs.
Toy companies especially are starting to catch on to the possibilities of 3-D printing, as they no longer need to set up $500,000 worth of tooling and months and months of engineering time to launch a new toy line. Now they could just sell a digital design file for $5 or $10 that took them maybe a month or two to design. This scenario is a rapid time to market at much lower risk.
“As 3-D printing tools become better and more automated, I can imagine a day where people buy just a circuit board for their new phone and print the rest at home or at the nearest 3-D printing center, customized to their needs.”
PwC: What are the prospects of printing integrated systems, particularly in the next five years?
Andrew Boggeri: I’m not truly sure. Optomec has a very interesting technology, but it’s not widely used yet. It would allow for a 3-D printed part with embedded electronic traces, but someone still needs to solder the parts. I think initially it will be confined to higherend products or high-performance products. It’s also a somewhat complicated system, so it won’t be on the desktop anytime soon. Will an integrated system printer be in every Joe’s house within five years? No, I don’t think so. But, I think in five years people will be aware of these things and trends.