Back in January, the Boston Consulting Group (BCG) published a study, “Surviving disruption in additive manufacturing” (access the study here).
What does BCG mean by “disruption”?
(1) From closed to open system
Currently, industry standard in additive manufacturing (AM) is an integrated, closed system: equipment providers sell both the printers and the materials. But in the future, according to BCG, this closed system will open up. AM materials will become its own category. And this category is highly attractive to big materials providers, because of the high margins.
(2) Materials providers will forward-integrate
AM requires deep knowledge about how materials on the one hand and printing technologies on the other hand fit together. Otherwise the quality of your products will suffer, or you will not even be able to make certain products at all.
Getting this material-printing-technology fit right means that you need to think in terms of material performance, not in terms of bulk. And this in turn probably means that as an AM material provider you will forward-integrate in your market.
If you put together these two developments, (1) moving closed to open ecosystem and (2) forward-integration of materials providers, this looks like a threat, or disruptive, to the current industry standard.
Can we see signs of additive manufacturing industry disruption?
What could such signs of AM industry disruption look like?
My article here is neither comprehensive, nor am I a subject matter expert. I just used the software we build here at Mergeflow, in order to address this question the best I could. Specifically, I decided to look at the following:
(1) Do venture capitalists’ investments reflect an “closed -> open system” change?
If AM really moves from closed to open system, this should reflect in venture investments. After all, VC investments are supposed to be bets on future, not on present or past, scenarios. For example, there might be investments in dedicated AM materials companies, rather than in the traditional “razor&blade” closed business model companies. Or there may be investments that try to hedge the question altogether, i.e. work out independently of whether “closed -> open” happens or not.
(2) What AM strategies do materials providers have?
Do materials providers forward-integrate? Are they getting into the AM “materials” space? If so, we should see evidence of such activity. For example, materials providers might patent attractive “white spots” in AM. This could be specific materials and their properties, or applications further downstream in the value chain. I will get back to this below.
Let’s start with the VCs.
Do VC investments reflect a “closed -> open” change?
I used Mergeflow’s venture investment analytics to discover and look at venture investments in the AM area in 2019 (click to enlarge):
And here are the investments, aggregated by topic (i.e. what do the companies do?):
It seems that a lot of money still goes into companies that make both, the materials and the printers. In terms of materials, composites do stick out a bit as their own category. But in composites, too, Impossible Objects, Fortify, and Markforged all seem to be running integrated models, providing materials and printers.
Then there is the “on-demand manufacturing platform” part, featuring companies such as Inventables, Xometry, Fictiv, 3D Hubs, and Fast Radius. Perhaps this is the “hedging part”, i.e. the investments where it probably doesn’t really matter if the “closed -> open” change happens or not. But since platforms have a tendency toward winner-takes-all (cf. the highly recommended book “Machine, Platform, Crowd”), it might not be too surprising if we see some M&A in this area further down the road.
OK, on to point (2), the materials providers.
What AM strategies do materials providers have?
There are many possible ways in which AM strategies could manifest themselves. For example, it could be investment activities such as corporate venturing. Or it could be patenting strategies.
Tech fields with patents but no published R&D can show unique market changing insights before they are realized in the market. After all, if someone holds patents in an area where nobody else seems to be doing anything, this might give them a particularly great advantage.
Here, I chose to focus on patenting strategies. In particular, I wanted to see if there are topics in AM where we see patents but no published R&D. This is because patents can indicate an intention. For example, a materials provider might file for patents in an area in order to ensure freedom to operate in that area. Of course, the incentive to do so is greatest if the company thinks they are working on a game-changing innovation. So this approach, “look for areas with patents but no published R&D”, can help you discover market-relevant, breakthrough innovations.
Using patenting and R&D data to infer strategy
Then, I searched Mergeflow for AM markets and applications, selected some of them, and grouped them like in the table below:
Note that in some areas (e.g. “3D bioprinting” or “3D printing medical devices”) Mergeflow identified more than one market estimate, with each estimate providing slightly different numbers.
Next, I put it all together, into a big search matrix: materials, methods, and applications.
I used Mergeflow’s Grid Search for the analysis. We used Grid Search for other topics as well, such as how to address climate change challenges with machine learning, or clinical trials at the intersection of nutrition, wearables, and lifestyle.
Notice that an alternative approach to Grid Search could be Mergeflow Teams. Mergeflow Teams supports collaborative tech discovery. The idea there is very similar to Grid Search. In Grid Search, you can look at the matrix, and identify results that are relevant to more than one of your topics. Similarly, Mergeflow Teams uses analytics that label your results with your team’s topics that match your findings. Basically, it is as if your team’s topics form the Grid Search rows and columns. Here, this would be one team member doing tech discovery for “polymers”, one for “metals”, one for “methods”, etc.. And then Mergeflow’s analytics does the row-column matching.
In order to identify areas with patents but without published R&D, I subtracted an R&D results matrix from a patents results matrix. The results are in the table below, which highlights in green the areas with patents but without published R&D:
Every row and every column represents a Mergeflow query. In the table we can see, for example, that the intersection of “polyetherimide” and “polycarbonate” is a patents-but-no-R&D area (marked by the red circle). I zoomed into this area as an example.
Turns out that some of the patents in this area were filed by SABIC:
2015: Additive manufacturing articles useful in aircraft, Patent at Espacenet.
2016: Polyetherimide composition and associated article and additive manufacturing method. Patent at Espacenet.
Why does this matter?
OK, nice, so now I found that SABIC holds some AM patents in an area with no published R&D so far. But why does this matter?
One way of asking the “so what?” question is to ask, what did SABIC do after filing these patents?
Inferring company strategy from tech and innovation data
I now used Mergeflow to find news and events that (1) happened after 2016, and (2) involved SABIC, additive manufacturing, polyetherimide, and polycarbonate. After all, if the 2015/2016 patents really mean something, we should see some new SABIC product or solution involving the patents (or the patents’ topics, strictly speaking) at some point further down the road.
Here’s some of what I found:
Global chemical giant SABIC details commitment to advancing additive manufacturing. Link to article
SABIC showcases new THERMOCOMP AM compounds and 3D printed yacht hull at formnext. Link to article
Here it is! The yacht! Thank you for reading all the way to here.
Airborne, Siemens and SABIC partner to mass produce thermoplastic composites. Link to article
So this does look like a rather coherent line from patents to products and solutions.
What have I learned?
My main lessons learned is that it’s worth applying analytics to combinations of data sets, not just to individual data silos:
- Markets provided some context (AM applications).
- Venture investments, combined with some digital business model basics, may indicate future areas of M&A interest (the AM marketplaces).
- Patents plus (or minus, rather) R&D data may indicate future areas of company activities (the SABIC example).
- Industry news helped me verify, or at least lend plausibility to, the “predictions” based on patenting and R&D activties.
Please note the quotation marks around the word “predictions” above. Making predictions is hard, if not impossible. Probably the best we can do is come up with plausible futures. This is how I mean “predictions” above. See also another article where we looked at making technology and innovation predictions.
I know that my SABIC example is a temporal sequence of events, and that this may, but does not have to, imply causal relations between events. And it is just one example. But it sure seems that “patents but no published R&D” might be an indicator of interest that is worth looking into a bit more closely and more broadly.