About Me

My photo
Ewloe, United Kingdom
Writing, tweeting, debating and occasionally getting a little over-excited about 3D Printing. But always aiming to keep it real!

Monday 31 July 2017

Hype, Hype Cycles and Applying Reason

Originally published at Disruptive Magazine. This is an unedited version .....

You know that feeling when you think you’re well and truly over something — believing it is in the past? That’s where I believed the inflated hysteria and hype around 3D printing had been banished. Turns out, not so much and I have been conflating belief with hope. Looking back, the mainstream media hype and hysteria around 3D printing peaked around 2012-2014, when many with longevity in the industry did their utmost to push back. Reality seemed to bite, though and things calmed down a great deal — even to the point of “negative hype.”

Over recent weeks and months, an increasing number of hyped headlines about 3D printing have crept back into focus. Some of them are simply clickbait — inflated headlines to get that reader to click through to somewhat more measured content. Others, however, are purely uninformed hype, backed by equally uninformed content. Both are irritating and frustrating, but the latter is more insidious in nature.

For the record, I don’t particularly like being negative, it is not generally my default setting. However, I’ve noticed myself turning rather shrew-like on Twitter recently as I ignore the voice telling me not to bite at some of the ridiculous 3D printing headlines that are proliferating across news sites and their social media channels once again.

The dilemma of “ignore the nonsense” versus “call this out for what it is” is a problem. Ignoring it often seems the easier, if lazier option and certainly involves lower blood pressure. But, for whatever reason, I feel a certain sense of responsibility with this stuff. Not calling it out just feels wrong. Experience and history shows us that these headlines might work to increase the readership numbers across a 24 hour cycle for the media outlets publishing them, but they also jeopardize the real progress of the 3D printing and additive manufacturing industry by raising expectations beyond the realms of reality, which in turn invariably leads to disappointment and ultimately increased cynicism and lower rates of adoption.

Indeed, at the recent International Conference on AM and 3D Printing, Phil Reeves of Stratasys Expert Services made just this point and cited how he believed that the historical mainstream media hype problem from 3-5 years ago damaged the additive manufacturing industry. He specifically made the point that the promised “revolution” was a misnomer. Dr Reeves presentation was centred around the existing barriers to adoption, however his inclusion of the inflated press coverage of 3D printing back then was telling. He went on to point to the much more “conservative” reality of today — one where some truly great applications and many more mundane but business-boosting applications of additive manufacturing are being adopted across a plethora of industrial sectors.

At the same conference, it was also both inspiring and sobering to witness the presentation of Pete Basiliere of analyst and research firm Gartner. This is the firm that developed and published the now famous Gartner Hype Cycle, which it defines as a  graphical depiction of a common pattern that arises with [a] new technology or other innovation.”
It was interesting to hear Pete, during his presentation, say that: “actually it’s more of a wave than a cycle.” I’ve pondered that a few times over the years. And in the context of this article, even while I may be missing the point, I can’t help but note that waves and cycles both have a tendency to reoccur!
In a neat coincidence, however, Pete’s presentation coincided with the publication of the 2017 3D Printing Hype Cycle.

Description: Macintosh HD:Users:rachelpark:Desktop:2017 Gartner Hype Cycle - Slide from Pete.png
Image Credit: Gartner.

What the 3D printing hype cycle does is identify and break down many different sub-sectors of the 3D printing and additive manufacturing industry and illustrate where Gartner believes they exist on the ‘cycle’ on their way to mainstream adoption. Gartner qualifies this within five time stages that run across the bottom of the graph, namely the:

• Innovation trigger (including early R&D; first start-ups and VC funding; 1G products, early adopters).
• Peak of inflated expectations (including mass media pick-ups & hype; supplier and funding proliferation; wider adoption, beginning of negative press coverage).
• Trough of disillusionment (including supplier consolidation and failures; 2nd & 3rd round VC funding; and less than 5% potential adoption).
• Slope of enlightenment (including development of methodologies and best practices; 3G products).
• Plateau of productivity (high growth adoption).

The pertinent arc over this slide is “mainstream adoption.” Pertinent because when you are embedded in a technology sector and overly familiar with both terminology and applications it can become easy to forget the “worldview” perspective. That said, any activity that involves prediction cannot be taken as gospel — it is not an exact science.

Scrutinizing the 2017 Gartner Hype Cycle, I mostly found myself nodding, but there were also a few surprises and a couple of entries that really took me aback.

Absolutely no surprise that “Consumer 3D printing” is still sliding down into the trough of disillusionment. I suspect it will remain there longer than Gartner’s predicted 5-10 years. With no specific mention of the maker community, I also wonder if this prolific user group of the desktop FFF machines fit into this category. This active and growing community group remains an underrated anomaly within the 3D printing industry.

Stereolithography traversing down into the same trough is a surprise. As the original additive process, and one that has been applied across many industry sectors I truly expected this to be well on its way up the enlightenment slope, but Gartner currently has it well behind material extrusion, material jetting and binder jetting.

Again, no surprise at all that 3D printing for prototyping has reached the plateau of productivity. I doubt anyone can seriously question that prototyping remains the 3D printing industry’s most widespread application, with correlating acceptance and increasing uptake.

But in a nice little plot twist (I thought), 3D printing of hearing devices is the only category Gartner places ahead of prototyping. It’s a production application, moreover it is predominantly a plastic production application, but even that is undergoing a transition — to metal.

To sum up — if you take nothing else away from this post, please just keep it real. The reality of the 3D printing and additive manufacturing industry in 2017 is exciting — remaining challenges included. The hype doesn’t help anybody.

Monday 17 July 2017

A Review of the International Conference on AM and 3D Printing

This conference, which took place at the Nottingham Belfry hotel and conference centre and is hosted by the University of Nottingham, once again, did not disappoint. The 2017 edition was the 12th annual event in this series and increased once again in both size and stature with in excess of 250 delegates and more than 30 exhibitors on site, according to the organizers. Following a well-trusted formula, the emphasis was well and truly on the provision of information — both through the high-level, in-depth conference presentations, and the intensive networking opportunities afforded during the conference days.

The quality of the conference programme was high and broken into two parts over three days. Day one took place on Tuesday 11th July and was run separately to the “main event” which took place over the next two days. The first day was dedicated to the “Industrial Realities of Additive Manufacturing” while the following days ran under the conference title.

In terms of the demographics of the attendees, it was unmistakably an educated and knowledgeable crowd, when it comes to AM. Prof Phill Dickens,’ who introduced the industrial day, took a straw poll that highlighted this nicely when the vast majority of hands raised indicated some involvement with AM in their work. There were a handful of delegates completely new to AM, seeking information for their business — and they picked a great place to start! Phill’s off-the-cuff poll also revealed a wide cross section of industries represented across the delegate base, including all the usual suspects such as aerospace, defence, automotive (elite and road cars) and medical plus a few others besides. As you might expect there was also a strong academic and research contingent.

Talking of ‘usual suspects,’ though, Phil Reeves, from Stratasys Expert Services, opened the Industrial Realities day with a presentation he entitled ‘Understanding the Production Economics – The Harsh Realities of 3D Printing.’  Some readers may understand the double entendre in the reference to ‘usual suspects’ – namely that Phil’s longevity in the additive industry, corresponding depth of knowledge and pragmatic approach sets him up better than most to deal with the ‘harsh realities.’ However, Phil’s presentation utilised the term “usual suspects” throughout to highlight the problems that industrial sectors face when implementing AM for part production, accompanied by suspect police line-up imagery on his slides.

Given that Phil only had 30 minutes, he did a stellar job of raising the issues and challenges that are often seen as barriers to adoption for many companies. After denouncing many of the press promises of AM and 3D printing, Phil highlighted how, in 2017 the reality is more …… conservative and additive technologies are not as widespread as we have been led to believe. The reason for this, Phil stated, comes down to five “usual suspects,” namely accuracy, build speed, part size, part cost and mechanical properties.

It’s hard to argue that these challenges are not still barriers to adoption. I still hear them cited by users of AM tech over and over again. Phil made an excellent point in his summation, however, that all five issues do not have to be solved all at same time. In terms of production applications with AM, the focus should always be on the application, and he believes more and more application specific hardware systems will emerge (think fuel nozzles, orthopaedic implants, hearing aids etc). Additive production systems developed and built for specific parts and components at higher volumes where the economics make sense. In this way, “the machines might cost $8 million, but it doesn’t matter if the value that comes off them justifies that investment.”

I do not think Phil is wrong here, it’s no secret that the huge multi-nationals are leading the charge with additive manufacturing for production applications; their deep pockets for hardware acquisition, integration capabilities and R&D make it a no brainer. That said, opportunities do exist for smaller and medium sized companies with additive manufacturing that should not necessarily be overlooked, but they do tend to involve more risk. This was the message from the presentation given by Sophie Jones, General Manager of AM consultancy firm Added Scientific. The presentation was centred around Sophie’s research, supported by Innovate UK, which included interviewing a number of smaller companies involved with AM in the UK. While it can be argued that this research has a regional bias, I think the barriers to adoption that Sophie identified for smaller organisations are universal.  

The first, and arguably the most significant challenge, cited by all the firms is access to finance. In this regard, the feedback illustrated how banks are reticent about funding for AM, largely because they do not understand the technology base and how to finance it. As a result, Sophie highlighted some hair raising examples of small, privately funded companies taking high personal risks to purchase of machines, it shouldn’t be this way, but if options are limited there is often no other route. Other challenges and barriers to adoption for SMEs raised in Sophie’s presentation included firms requiring back up revenue streams to support AM activities; the issue of global supply chains, while sales were largely domestic; the need to educate customers; the AM skills shortage; and, last but not least, industry accreditation – ISO accreditation is vital, as most customers demand it, particularly those working within highly regulated industries.

Looking to other highlights from across the three days, and one that stood out was an evening meeting organised as an extra-curricular activity by Sophie Jones, by invitation only, for women in AM. It was well attended, by 20 women, a testament to the noticeable increase in women working in this field. However, the fact that this is even an issue that needs highlighting, and that the percentages overall are still low, means there is still much to be done.

Highlights from the conference proper were many and varied. The research into a new additive process being undertaken at Lawrence Livermore National Laboratory (LLNL) — presented by Maxim Shusteff — did stand out. This is a new, faster photopolymer process, called Fast Volumetric Fabrication. The science presented lost me more than once, but a video clip illustrating part formation in the resin bath — in a couple of seconds — blew my mind. I was not alone. There are no layers involved in this, the process is enabled by a truly three-dimensional holographic light source. As always, with new processes however,  Maxim qualified his excitement about this new process (increased speed, no free surface, no substrate, and more predictable/traceable process models) with: “Holography is interesting, and we’ve shown it’s possible, but it has limitations.”  

And the holographic process is not the end of the story either, Maxim also provided a sneak peak at another new process under R&D at LLNL, this one called Tomographic Volumetric 3D Fabrication, which shows promise in terms of eliminating geometric limitations. The research on this process is due to be presented at the upcoming SFF event.

As an aside, after this presentation, someone next to me commented: “makes the ‘Star Trek Replicator’ analogy seem possible.” I disagreed, in the strongest possible terms. And for the record, “Tea, Earl Grey, Hot” — the American script-writers of Star Trek aside, who on this planet needs to qualify that Earl Grey needs to be hot??

In terms of newer, commercial (or nearly commercial) processes there were some very insightful ‘X’ presentations from Neil Hopkinson and Dror Danai of Xaar and XJET respectively. The high speed sintering process itself has been well documented and Neil, as ever, delivered an accomplished presentation; however the Xaar business model both with this process as a service and internal tool continues to intrigue and will, I suspect prove disruptive. Dror’s presentation provided some real insight into the almost-ready-for-commercialisation nano particle jetting (NPJ) process with both metal and ceramic materials. The R&D model at XJET is beyond impressive, I discovered when I talked directly with Dror at the event. The “magic” behind this process lies in controlling the delivery of the nano particles in a proprietary dispersion material. The enabling tool is the proprietary inkjet head, and the temperatures it can withstand. Not a dissimilar narrative to Xaar, actually. Moreover, the anecdote I heard more than five years ago, about “inkjet being the future of additive manufacturing” kept coming back to me during the conference.

In terms of advanced applications highlighted at the conference, delegates were enlightened on some of the intricacies of producing parts for performance bikes for the UK Olympic and Tour de France teams (METRON), Cars (BMW), electronic products (Texas Instruments), and hearing aids (Sonova). There was also a long-term AM vision presentation from Airbus, but this was generic in nature, with no specific applications referenced. While none of these applications are wholly novel in terms of the sectors, the Sonova presentation in particular highlighted the over-arching narrative of progress with AM, and what can be achieved now, compared with when it was first implemented. Sonova, for instance, via its various brands has been producing millions of small production hearing aid parts with plastic AM since 2007. Ten years later, in 2017 the company has transitioned to metal AM – with biocompatible materials and improved functionality — at the same volumes.  This is a really big deal — for the company, for the technology and a great marker of progress for the additive industry as a whole.

I mentioned Xaar’s business model above, and I will be digging deeper into this via a new source just as soon as I get clearance. However, another company that revealed a very interesting business model with AM is Johnson Matthey (JM), a world leading catalyst manufacturer, with core competencies of developing catalytic materials, coating, powder production and ceramics. Funnily enough, the presenter, Samanth O’Callaghan joined Johnson Matthey from Xaar two and half years ago. Regardless, JM after initial research into various AM processes in 2009, initiated a very specific, application based solution for AM by developing a binder inkjet ceramic AM process. It’s another interesting business model, once again based on user evolution, whereby JM also developed and uses its own ceramic materials — not a surprise really, considering the company’s expertise.

Of note during Samantha’s presentation was the positive qualification for using this process, namely the scalability of binder inkjet technology and the facts that it is “faster and cheaper at scale.” She highlighted the significant post-processing requirements, and how porous parts are; stressing that this was actually an advantage for the JM application. Today, JM is able to make a better product, with greater sustainability and cheaper, with its ceramic AM process. The company is in the process of establishing its pilot plant, which will be completed in a few months, and will be manufacturing at scale; “tonnes per year,” according to Samantha, with automated bespoke material handling and an integrated end to end solution.  

The scope of the programme was wide, and the depth of the presentations, individually was so impressive and informative, that it is impossible to do it full justice in one round-up article. However across the three days, four themes kept recurring — in presentations and conversations — notably the skills shortage around AM; Funding; AM integration into factory workflows, both digitally and physically; and how this is often resulting in a hybrid (subtractive and additive) workflows.

The content from the three days has given me, and I suspect all the delegates, much to think about. I certainly have plenty to follow up on and write about for the next few weeks (months?). 

Tuesday 4 July 2017

The AM Ancillary Problem

OpEd for Disruptive Insight, Issue 6 (unedited)

There is a problem around the Additive Manufacturing (AM) and 3D printing industry that never quite gets enough oxygen but is never far from the surface. It certainly is not going to go away any time soon.

I would be the last person (or at least one of the last people) to dispute the amazing capabilities of AM and 3D printing hardware systems used for industrial applications. I never get bored of watching the faces of people seeing a complex part off an AM machine for the first time, and the incredulous looks increasing as the process is explained to them. And yes, while more infrequent these days, it still happens. (But that’s a story for another day.)

The ability to build complex parts, in one piece, and the advantages this brings of increased strength, lighter weight, reduced material consumption and assembly component consolidation for an increasing range of applications are all well documented and justifiable drivers for this technology group.

However, oftentimes, the focus for these advantages is singularly the additive hardware that build the parts layer by layer. This is understandable up to a point — it’s the enabling process where the advantages listed above materialise — but in reality, for new (and even regular) users, it does not convey the full picture of what is required to get that part “off the machine.”

AM hardware systems are part of an extensive ecosystem of technologies that enbable it, both pre- and post-process. These days there is greater emphasis being placed on the pre-processing discipline of Design for AM (DfAM), file preparation and file format in terms of technology development, debate and awareness. In the previous issue of Disruptive Insight Kruno Knezic addressed some the the issues around design software and skills and highlights some of the progress.

Where there is still shade is around the in-process and post-processing extras — specifically the ancillary hardware that is often critical to an additive process but gets overlooked in terms of its contribution to the end result, the time it adds to the actual build time, the space it consumes and, most notably, the cost it adds to the purchase price — rarely talked about.

This is not actually a new problem born of the many new processes and machines now available on the market. Historically, Stereolithography systems have always required curing ovens, often larger than the AM machine itself. Similarly, the Laser Sintering machines of the 1990’s needed powder handling /removal systems and recycling hoppers and sieves. Not to mention the Personal Protective Equipment (PPE) for machine operators. The post-processing options for plastic sintered parts back then, also invariably required infiltration operations, as well as finishing processes, particularly if aesthetics were important alongside the strength advantages that laser sintering offered. The Fused Deposition Modelling (FDM) process also required finishing processes to eliminate the very obvious stepping effect of the process during its early years —baths of chemicals, and later water baths for water soluble supports  were common requirements, not to mention the endless sanding .

As we head into the final quarter of this current decade, additive processes and the materials used to build parts are considerably more advanced than they were in the 1990’s. I’m stating the obvious now, but the materials palette is much broader; resolution, accuracy and repeatability are consistently meeting the industrial requirements for critical prototyping, tooling and some production applications. What is still frequently overlooked, however, is that these achievements often require secondary processes — that barely ever get a mention. Certainly not in the marketing materials or the sales pitches. As stated above these secondary processes require considerable investment, financially (of course) that can double the price of the actual 3D printer; but also time and floor space.

It is noticeable that the term “post-processing” is often used interchangeably with “finishing” — this is somewhat of a misnomer, actually. With many processes there are a series of essential post-processing steps prior to the 3D printed part finishing stages.

Today, most resin polymer additive processes still require oven curing. Laser sintering and laser melting additive processes still require powder handling equipment. On an industrial scale this can be twice or even three times the footprint of the additive hardware itself. For metal parts, removal from the build plate is a particularly undesirable part of post-processing, and that’s before even starting on the support structures. Moreover, the PPE equipment for metal powder processes is more essential than ever.

But even consider the new, more accessible plastic powder bed hardware from Formlabs. Launched this month, the Fuse 1 system is an industrial desktop machine that offers, proportionally, the power and material advantages of a powder bed fusion system in a compact, more affordable format than alternatives on the market. There is no mistaking that it is a welcome addition to the market — accessibility is one of the major keys that will continue to drive uptake and adoption. But the eye-catching headline price of this system is $9999 / €12098. Or, if you look at the website, it “starts at” those prices. That’s because, when you look at the technical specifications, the actual printer IS that price, but a “complete SLS solution” requires considerable ancillary equipment. Indeed, the complete solution includes the printer, a post-processing station and intuitive software for setting up and managing prints — and costs twice the “starting” price, at $19,999.

It was actually a conversation about the Fuse 1, in this regard, that triggered this article. And it made me think on how often the excitement of a new launch often overlooks the ancillaries issue.

HP did something similar when they launched too, if you remember — launching the Multi Jet Fusion additive system months ahead of the HP Jet Fusion Processing Station, required to again provide the complete solution. The ancillary equipment required for MJF parts post build easily doubles the footprint of the 3D printer, and it’s not that small to begin with, and it adds considerably to the capital expenditure cost.

So it behoves to look around at some of the other new(er) additive system offerings with ancillaries in mind. And actually it becomes obvious that some vendors are aware of the issues with ancillaries and while they are not talking about them, there is evidence that they are attempting to overcome some of them with their tech development. That said, the familiar need to make a trade-off, depending on application requirements, process requirements and budget etc, remains. As it has always been.

Carbon’s CLIP process, launched originally on the M1 3D printer, is a resin polymer process that produces parts that need to be cleaned and cured. The significantly faster speed of the printing process itself, is offset somewhat by these post-processing steps. However, Carbon’s business model goes some way to mitigate this. First, the company’s business model is subscription based, meaning that the the costs include all the ancillary, post-processing equipment (as well as the software updates). Moreover, the 2nd generation of the technology — in the SpeedCell format, with the M2 printer — includes increasing automation for the cleaning phase. It’s all moving in the right direction.

Desktop Metal  is another company addressing some of the ancillary issues head on. DM launched its new powder metal 3D printers with full visibility of the ancillary furnace that is required post-build. In terms of part removal and support removal, the company has addressed this within its technology development to reduce the pain-staking and time required for these post-processing steps. Another trade off.

And then there is Rize: post-processing, or rather the lack thereof, is one of this company’s USP’s. I gave them a bit of a hard time a couple of months back, for using a tagline of “zero post-processing.” At the time, it was about sensationalized marketing, which I posited was actually making potential users more cynical than they might otherwise be. The point I was making is that amid its competition, the truth (ie minimal post-processing, relatively speaking) will get plenty of attention. It’s actually a very big deal when lined up against its cohorts.

The point here though is not about promoting any one technology over another. There is very little point in doing that these days because process selection is (and should be) application driven. The point is to raise awareness of the ancillaries required to make 3D printers work — at any level. It’s kind of like the insurance equivalent of the small print, so — ask the questions, and if possible, talk to experienced users, that’s always where you’ll find unequivocal truth.