Your web-browser is very outdated, and as such, this website may not display properly. Please consider upgrading to a modern, faster and more secure browser. Click here to do so.
While mainstream adoption is still many years away, 3D printing is already common in certain niche applications. The key success drivers to adoption of 3D printing for a particular application are:
Low Quantities – 3D printing technology is typically only economical for low production quantities. As quantities increase, its higher production costs make it uncompetitive.
High Willingness-to-Pay – Since production and material costs are significantly higher with 3D printing; industries that are extremely cost sensitive are not good candidates for adoption.
High Complexity – Products demanding complex forms help justify the increased cost of 3D printing. Traditionally, complexity can require multiple production technologies and assembly steps. As described above, complexity is ‘free’ in 3D printing.
Supply Chain Impact – The unique tooling and setup costs of 3D printing mean that it can be quite disruptive in small niches of the supply chain. Industries and applications that have high supply chain costs relative to products costs are good candidates for adoption.
Data Availability - All 3D printers need computer data to operate. Certain industries and applications have an advantage in having a wealth of data available so that they reduce the content creation barrier to adoption.
The following industries and applications can be broken down along these factors. Note that most applications require that several factors be favorable for adoption.
Medical Devices -Medical devices have been using 3D printing technology for quite some time. There are several factors for this. All custom medical devices have production values of one. Products are purchased on performance vs. cost. Complexity is high for prosthetics, etc. And, the recent advantages in scanning technologies mean that there is a wealth of digital data available. An example application is Invisalign. Invisalign 3D prints series of orthodontic correction devices for their customers from their dental scans.
Aerospace - As with many new technologies, aerospace is one of the first major industries to adopt. Performance requirements are high and production volumes are much lower as compared with consumer devices. There is also a high willingness-to-pay. Finally, since programs can last decades, keeping the required spare parts on hand is very difficult and costly. Examples of applications of 3D printing in aerospace include instrument panels made by RC Allen. General Electric Aviation recently acquired one of the major metal 3D printing service providers, Morris Technologies.
Niche Markets - Traditionally, when small companies have ideas for physical products they often cannot execute on their ideas because of the large fixed costs associated with having something produced. These barriers are now removed. Kappius components are a great example. Kappius makes very racing bicycle components for a particular style of competition. Due to volumes, high customer willingness to pay, and high product complexity, 3D printing was the best technology for production.
Spare Parts - While most businesses try to avoid inventory, inventory is the business in the spare parts industry. A huge variety of parts have to be held for many years. Rather than buying and holding, 3D printing could print spare parts on demand. Volumes are low enough to remain cost competitive. And customers typical have a very high willingness-to-pay since the parts may well be critical for larger equipment. NASA has discussed using a 3D printer in space to produce parts when needed. And newer companies, such as the Swedish music products business Teenage Engineering, are utilizing 3D printing to eliminate that portion of their supply chain. Teenage Engineering now posts the CAD files for the spare parts for free and will tell users where to have them printed. (3Ders.org)
Mass – Mass Customization - Shapeways is a New York based venture funded company that creates a market for consumer to consumer sales. Individuals can upload their own designs for anyone to purchase. Shapeways performs the 3D printing for all designs sorted through their site. Additionally, Shapeways has solved the ‘CAD data issues’ as users provide the necessary CAD designs. Shapeways has received $47MM in venture funding to date. (Tech Crunch)
A summary of these factors applied across industries is below:
Industrial equipment is normally a pretty straightforward affair. ABB and Kuka make industrial robots and sell them to manufacturers including the auto industry. ASML makes computer-chip-manufacturing equipment and sells to Intel and other chip manufacturers.
But something is different in 3D printing. Stratasys and 3D Systems are the two biggest equipment manufacturers. However, they are also two of the biggest 3D printing service providers.
Stratasys has its RedEye Direct service while 3D Systems has its QuickParts service (note that Quickparts actually offers many additional services such as molding a machining).
3D Systems has made huge investments in its service offerings having acquired 15 providers since October 2009. Part production revenues have grown from $18.3MM in 2010 to $79.2MM in 2012 (note that the acquisitions mentioned above likely accounted for much of this growth). Production is a key part of their long term strategy and accounts for a sizeable portion of revenue.
Direct Part Production is a much smaller, but still important, portion of Stratasys’s business. 2012 service revenue (all service) grew by $7.1MM, or 24.9%. This also reflects $3.0MM in service revenue from Objet. Note that precise part production revenues are a bit more difficult to tease out in the case of Stratasys since services are all lumped together.
ExOne, another industrial 3D Printer manufacturer, has also recently announced a network of print facilities.
The fact that these companies are attempting to grow their production businesses is quite interesting. It speaks to the power of 3D printing. It is a technology that can readily serve a wide range of applications across multiple industries.
It makes sense in some ways. They actually have tremendous advantages over any other service bureaus. After all, everyone else in the playing field has to pay market rates for the two biggest cost drivers: machines and materials. They get both ‘at cost’.
However, I would tend to think that ultimately Stratasys and 3D Systems have more interest in making 3D printers than being service bureaus. After all, this is where their core ‘defensible’ competency lies. And where they have a sizeable lead over the competition.
To promote broader adoption and expand the installation base, we hope the large equipment manufacturers adopt the “let a thousand flowers” bloom approach.
3D Printing can be used for production.But to make it a fair fight, you have to pick the right combination of geometry, margins, complexity, and quantity.The spare parts industry is touted as an area where it may first find broad applicability.We decided to take a quick look.
At a high level, this makes sense.Quantities are low compared to new products.And margins are high.People will pay a lot more for a component when it is no longer widely available and a key part in operations. The fact that CAD data exists for many parts also lowers the barrier to adoption.
For comparison, we consider two different models for spare parts fulfillment:
This can get quite tricky as there are many variables involved.We analyzed these two models under the following conditions:
The Spare Parts Warehouse –While most production facilities look to minimize inventory, the business of spare parts is inventory itself.On the one hand, they are very efficient at dealing with inventory and face a lower risk of obsolescence than most companies.On the other hand ….. inventory is still inventory and the spare parts organization must balance service level vs. costs.
Cost of material is assumed to be $0.25/cm^3.Fixed ordering cost is $200 a small production run must be setup for each restocking.As a result of the high relative ordering costs, inventory levels are high compared to most businesses.This makes sense intuitively.
The key to remember here is that spare parts must be held for each and every different product variants.For low demand volumes, this adds up quickly over 1000’s of separate parts.
On Demand Spare Parts with 3D Printing - Using ‘Digital Manufacturing’ or any other ‘tool-free’ production, the dynamics change significantly.Now, ‘uncertainty’ does not need to be calculated for every product.Rather, demand can be aggregated across *materials*.This is a very different type of aggregation and postponement than we normally see.
This dramatically reduces the amount of ‘inventory’ being held.Inventory is now powder or another additive material is raw form vs. pre-produced units of inventory.
For the sake of comparison, material costs are estimated to be double that of traditional manufacture.And fixed ordering cost is assumed to be $100 for new material.
Assumptions – The following is a list of assumptions made in the model.There are some pretty big simplifications for tractability.However, none of them fundamentally change the dynamics.
Results – Below is a summary of the results for differing levels of demand.Weekly demand is varied for all 5000 products.For each level of demand (1-5 units weekly) the cost breakdown of a traditional warehouse is on the left and the 3D printed costs are on the right.
From the chart, we can see the following:
Lessons – This simple model highlights some important trends:
People should be very excited at the prospects of 3D printing.But, they should also remember to think about where and how it will start to be used in practice. Hopefully this provides some insights into these problems.
We are happy to provide more detail on the model or answer any questions you may have.
Or …. “Additive Manufacturing as a Real Option”. Or … “Perhaps the only blog thus far with options pricing and 3D printing”.
In a previous post, we compared the costs of production using 3D printing vs. low volume injection molding. The title of the post was “cost = f(size, quantity, technology) + a whole lot more”.
It got some good reaction. A critique from Shapeways itself here. Even a retweet of this critique by Chris Anderson. The point of the blog is that at some combination of those variables, 3D printing is the right technology choice. And, at some point (even with density discounts), it is not.
In his comments, Duann Scott(@DuannS) mentioned the idea of being able to subsequently alter design as a benefit of using 3D Printing for production.
I wholeheartedly agree and decided to dig in a bit with some numbers.
At the highest level, using 3D printing for early production can be a hedge, just like an option. You are paying a premium to have the right to invest later. For production, this means eating higher production costs in order to:
a) avoid a big flop (i.e. mass producing a loser)
b) permit later design improvements
This idea is very powerful.
You can get as complex as you like analyzing this. Stochastic optimization models or Black Scholes pricing would be advanced methods. We will go a simple route with a great little Excel add-in called TreePlan. The idea is the same, but this simplistic framework helps to demonstrate the concepts.
To avoid similar pricing criticism, I will use a some data directly from Stratasys/RedEye Direct. The data is for production of a medical device casing and was re-entered into Excel for comparison. The original presentation was made at a Minitek conference a few years back where they compare injection molding costs vs. low volume injection molding. They apparently have no problem comparing apples and oranges ;)
Let’s say you are developing this product. You estimate demand to be 250 units. But this is uncertain. For simplicity, you assume that there is a 50% chance of selling all 250. And a 50% chance of bombing and only selling 50 units. Your profit per unit (allocated to the casing) is $140.
You could approach this uncertainty several different ways. One way to play out these different options is using a decision model. It lays out the decisions available and also the uncertainty in different scenarios. The scenarios play out from left to right. Ultimately there are 6 different outcomes in this case. Each has a final profit assosciated with it.
Initial Decision - To 3D Print or Not to 3D Print? - You have a product. You have some estimate of demand. You can either just go ahead with full production or test the waters with a small first run.
Phase 2 - Stay the Course or Give Up? - Let’s say that you went the 3D printing route and you are still uncertain of demand. You now have options.
The nice part about this approach is that the model looks across the outcomes and their relatively outcomes. In this case, it predicts about a $4K profit starting with 3D Printing and a $3.5K profit if you mold at first.
Note that this model is very sensitive to the quantities modeled. Changing the first run by 10 units can really change the results. This is due to the fact that at these low volumes, we are operating in the highly non-linear portions of the price curves.
There are a ton of improvements you can make to the model. More detail on design and redesign costs, improved sales projection distributions, etc, etc, etc.
But, the idea is still there. 3D printing is giving you an option to hedge your bets. You reserve the option to evaluate product success and redesign if needed.
If you are a small producer, you can do this by feel, basic analysis, and even opening a shop on Shapeways and other marketplaces.
If you are making large quantities or high value goods, perhaps this level of rigor makes sense. You can actually get a lot more precise and figure the optimal production quantities at each step.
The same tradeoffs exist with other technologies as well. Custom PCB manufacture, rapid tooling and other low volume production techniques give you the same options before large scale, outsourced production.
Email firstname.lastname@example.org if you have questions or would like more details on the analysis.
There have been about a million stories about the 3D printed Nike Cleat. Part of me is loathe to post another. But, I really do think that this project highlights some important things about how companies should and will utilize 3D printing.
Don’t replace …. reinvent
3D printing is a long way from being economical for most products. It is quite simply not feasible to replace mass production techniques. So, don’t play that game. Don’t replace existing products. Instead, create new ones which necessitate(justify) the technology.
“SLS technology has revolutionized the way we design cleat plates – even beyond football – and gives Nike the ability to create solutions that were not possible within the constraints of traditional manufacturing processes,” said Shane Kohatsu, Nike’s director of footwear innovation. (SmartPlanet.com)
Or, as stated in the 2012 Wohlers Report: ”Perhaps the most significant barrier to realizing new applications and powerful benefits is our reluctance to accept new ideas and methods. Established processes and procedures are difficult to displace.”
Parts …. no Products
There is a vision of 3D printers spitting out finished products, massively disrupting the supply chain. This day may come, but it is a long ways off. With few exceptions, such as Optomec and Objet along with a few others, 3D printers print in single materials.
Take a look around you and see how many items are made of a single material. Not many. And certainly not many you would pay a premium for. The possible exception being jewelry, cases, and home goods.
It is certainly exciting to see these objects being produced on a small scale. However, this does not represent a seismic shift in global supply chains.
In the Nike cleat, we see a better interim fit for 3D printing. The 3D printed cleat base is part of a larger product. It is used as a key differentiator and thus justifies the increased cost.
A model for growing the use of 3D Printing
Before 3D printers are spewing out finished goods, many companies will leverage the technology in this same way. They will use the unique design and manufacturing capabilities of 3D printing to advance larger products and brands.
In the previous post, we took a quick look at the costs of 3D printing vs. Injection Molding. Duann Scott at Shapeways cited the blog and made some great observations. The broader point is that both technologies have a place. And for a single part, there is a crossover point when tooling investment makes sense.
At the right volumes, 3D printing can be a replacement for traditional manufacturing for a given part. However, the true potential lies in unlocking the unique material, geometric, and manufacturing capabilities of additive manufacturing.
One way to look at design quality is through the complexity lens. Complexity is a bit of a tricky animal. It does not translate with mathematical precision to product quality, cost, or reliability. However, it has a strong relationship to all of them. Simple products are better products.
However, this makes the comparison of technologies even more difficult, since we are now comparing both manufacture and design. However, since we like to quantify things at 3Sourceful, let’s take a SWAG.
For a quick look, we borrow a framework from Xiki and Filippo Salustri (link below). There is a whole host of other metrics that you could use alternatively. This metric is empirical:
C = 1/f * (Np x Nt x Ni) ^ 1/3
We re-examine the two examples covered in the case,the strap mechanism for translational loading. And a flange mechanism with translational loading with rotational freedom. The analysis is largely the same. However, 3D printing design and fabrication allows the following changes:
Assembly / Strap Mechanism - Assume that the # of parts cannot be reduced. If, for example, the entire assembly is built using SLS, the pin and straps can now be the same material. However, overall product complexity is not reduced that dramatically as the number of separate parts and function remains largely the same. Overall complexity drops from 3.42 to 2.71. A good reduction, not not earth shattering.
Flange - The flange assembly offers a chance for dramatic reduction in complexity. Assume here that the entire flange is built from a single material. The number of parts is reduced dramatically. The number of interfaces drops as a result. And the number of features drops since less steps (like drilling) are required. Here, we see the overall product complexity drop from 3.66 to 0.66. This is huge.
There are additional factors to consider. How do we treat ‘digital materials’ such as those produced by Objet machines. With digital materials, properties can vary continuously. Are they separate or single materials?
And, the total cost analysis becomes quite difficult. There will be savings due to assembly complexity (features, interfaces). However, the manufacturing cost itself will likely rise. Examining all the tradeoffs here is quite difficult.
But, the overall message is clear. 3D printing, while requiring more forethought in design and costing, can offer dramatic opportunities to reduce product complexity. Thanks again to Filipo and xiki for a great case study to start from.
Was recently looking at costs for producing two different items. We know that the true strength of 3D printing is in customization and geometric complexity. But, still interesting to see how it compares in production costs for small quantities.
In this case, we prices out two different parts. One, a very small bracket (~1cm^3) and one a larger jig (~50 cm^3). To compare, we obtained quotes from Shapeways and Protomold. And for simplicity, we just assume the cheapest material from each.
We then plotted out the *total* cost of production for different quantities. As we would expect, the tooling costs of the molds resulted in 3D printing being cheaper at lower quantities in both cases.
But, in the case of the larger part, the cost of the 3D printing material meant that over 100 units, Protomold became the cheaper solution. Where, for the smaller part, 3D printing was cost effective over 1000 units.
This is a very simple example, but highlights some important points:
And, most importantly, it is crucial to explore materials, vendors and technologies when pricing out a part. The savings can be significant.
3Sourceful is a new service intended to help businesses make the leap from prototyping to production.
We are just getting off the ground. Talking to companies about how they want to utilize digital manufacturing.
Our vision is a marketplace where companies and designers can get multiple manufacturing bids for their CAD models. Because no single technology or company fits all.
3D printing can make sense for small order sizes and high product complexity. But rapid tooling / injection molding may be better for larger volumes and finishes.
Just as CAD helps explore the design space, 3Sourceful will allow the exploration of the production space.
We are now open for registration. If you are a designer, tell us what you need made. If you are a manufacturer, tell us what you can make.