Results from Design for Additive Manufacturing

A month ago we wrote an entry that showed an exercise to incorporate greater complexity into 3D printed products. The article makes a short review of the role of complexity in product architecture and how it influences our manufacturing costs and the way we build businesses around a product. Then, we continued to explain how the main advantage of 3D printing is the available free complexity and how it can be harnessed through a concurrent design process. Through the process proposed by Remi Ponche and his team of researchers, we can incorporate the printer as a connector of functional components despite its complexity. We finally concluded with an example of the implemented process in a little workshop carried out in the Ballerup library located in Copenhagen Denmark. In this example, we implemented a “forced analogy” ideation method to connect concepts that were very difficult to relate. Following the process proposed by Dr Ponche et al. we used the available tools in the software and printer to create a product that exploited the advantages of 3D printing.

Untitled-4
The process incorporated tools that are readily available for hobbyists and an FDM MakerBot 3D printer.

The final product was being printed by the time the post was published. Therefore, we wanted to show the final result to wrap up the complete process. This product, which we call “Cyggle” ( a mixture of danish words CYKLE & Gløgg) took 80 hours for printing and 750 gms of PLA platic, a bit too much, but a good start for a first product iteration.

27145048_10155847333020435_966345130_o
The final “Cyggle” product.

According to the process, the Cyggle was produced to damp uncontrolled bicycle handlebar movements when we ride home drunk from a party. The volume was optimized to transfer the force of the handlebar to the frame through a “rib” volume that goes around (that is why it has a round shape). Additionally, we included a beer holder just in case party is not over.

27140609_10155847332860435_793696006_o
From the back, we can see the cup holder and the clamp used to fix the product to the bike frame.

The 1.5 mm fins that make our structure flexible were printed without problems. As we can see in the images above, the printing orientation aligns with the direction of the force between the handlebar and the frame. Accordingly, a minimal amount of support material was needed. Over the final bicycle, the product fitted correctly.  The clamp mechanism should be redesigned since the current one does not allow us to remove the Cyggle after the first installation. The cup holder dimensions work perfectly but could use some support for “tall boy” cans and bottles.

The flexibility of the fins works just as simulated creating more resistance as they touch each other. Trials of fatigue could be held with this prototype but there is no visible sign of extreme exhaustion. Nevertheless, the contact surface can also be optimized to damp the shear forces when in contact with the handlebar.

Overall the process demonstrates that complexity in product architecture is achievable with 3D printing regardless the layering process. The fact that this example makes use of a regular MakerBot FDM printer proves that the most simple 3D printers present in local FABLABS are enough to bring up complex ideas that can give birth to innovative concepts and businesses.

Connecting complex components with Design for additive manufacturing: an example…

If you have already read some of the posts in this blog, you can tell that one of the main topics of this research project is the relationship between the product architecture and 3D printing. I refer here to product architecture as the product’s internal collection of components and relationships between them. Product architecture has been proven to affect the businesses ability to change, evolve, and create new products and services. One way it influences the development of businesses is through its level of complexity. Complexity is the property of systems that show us how intricate is a system inside. Complexity increases when one of two things happen, either the number of components inside the system raises or the number of relationships between the existing components also increases. This complexity of product architecture is reflected in product costs. While we usually relate costs to the volume of raw material used in fabrication, complexity adds costs by increasing the number of operations needed to transform the raw material. The more complex a product, the more operations we must perform in order to build it, and as a result, costs increase. This means that manufacturing a cube only requires enough operations to create six flat surfaces while creating an action figure, requires enough operations to sculpt more and more difficult surfaces. Complex architectures do not only have more features or faces to manufacture but also interfaces between those features and external components that need fabrication. As a result, products with more complex architectures are costly to produce, they need more processes, and specialized equipment and personnel. However, with the features that allow 3D printers to build objects in layers, this complexity cost is no longer relevant. 3D printing fabrication relies on computer controlled tools that selectively deposit or fix material regardless the feature complexity. As a result, product architectures with greater complexity have the same manufacturing cost in 3D printing processes as long as they have a similar volume. Therefore, manufacturing our products using 3D printing (or Additive Manufacturing) allows us to increase the number of components of our products (our product architecture) and the number of connections between them.

cube complexity
In this example is quite clear how the Polygon T-Rex has more surfaces which would require more manufacturing operations than the simplest cube to the right. Nevertheless, as long as they have the same volume, the costs in 3D printing will be the same. ( T- Rex from https://www.cgtrader.com/3d-models/animals/dinosaur/head-t-rex-dinosaur)

Contrary to other features such as custom or remote fabrication, there is a very good reason to focus on the creation of complex systems. Human cognition classifies information in boxes in order to cope with the environment around us. From the very beginning of human evolution, we started designating spaces, objects, agents, etc. This feature of our minds is commonly referred as Bounded Rationality and helps us make the world more predictable. Every time we see a dog, we already know what the “DOG” box contains, how it looks and behaves, and how can we interact with it. We do not have to spend time observing and experimenting with the dog until we get to know it all over again, we have all agreed on what the “DOG” box means. Likewise, when we create a new product, we create boxes for its product architecture. Systems, subsystems, components, and subcomponents are placed individually in boxes inside our minds. Those boxes include everything we know about them, and all the knowledge needed to fabricate them. Accordingly, complex product architectures include more and bigger boxes which means more knowledge invested in the product. Every business depends on differentiation from other businesses for success, therefore, having more knowledge than the others implies more differentiation and better sales. The fact that 3D printers allow us to create more complex product architectures without increasing our investment is an opportunity to create better differentiated and innovative products with more knowledge inside them.

giphy
The features available in 3D printing allow the creation of complex product architectures just like this robot created at MIT’s Computer Science and Artificial Intelligence Laboratory(CSAIL) in 2016.

Nevertheless, as the use of 3D printing as a manufacturing process becomes popular, it is evident that most of the products created using this process are not new and complex but products that have been designed for other processes and are now produced with 3D printers. Some researchers suggest that the main challenge of 3D printing implementation is the way our understanding of how design and fabrication work together. In a very interesting research paper, Dr Remi Ponche and a team of researchers show that the traditional way design and manufacturing work together is essentially decoupled. In other words; we are used to design first, and develop the manufacturing process later. As designers we are accustomed to create a geometry, considering an idea of what can be fabricated, and later select a manufacturing process. We adapt the design details locally, in order to make it feasible under the given budget. This way of designing makes sense if we use traditional (subtractive) fabrication methods because adding more complexity increases costs. Altogether, traditional design looks for lower manufacturing costs.  Thus, when we see products that do not exploit the features of 3D printing, we can say that the design process used to create them does not consider the lack of “complexity cost” that is available through 3D printing. Designing products in a traditional way only narrows the available complexity and as a result the available differentiation in a product architecture.

Dr Ponche and his team suggest a global approach to design for additive manufacturing, where we can design using the 3D printing process as a tool to create complex architectures. Contrary to the traditional approach, the global perspective exploits the complexity freedom in the 3D printing process by defining the final geometry of the product from the manufacturing process and the final product functions. They propose a design process that can be used in different 3D printers regardless the technology they use to bind layer by layer.

Before starting we must be sure of our requirements just as in every design process. We must be sure of the problem that we are trying to solve. What is our product’s final purpose? Does it have mechanical requirements? Where and when is it going to be used?

  1. Once considering the design requirements, Ponche et al. suggest a geometrical analysis considering the dimensions of the printable volume and the volume needed to solve our product. Note that printers these days have small printable volumes around 20x20x20 cm. This does not mean that we are limited to build small components, on the contrary, it only tells us that in order to create bigger products we must build them through the fabrication of small modules or use existing materials that help us expand the dimensions of our solution (the use of poles is a very good example, poles give us volumes that we can use as a skeleton to place printed modules around)  
  2. In the second step, we determine the shape of “functional volumes”. These volumes are the components that perform a function in our product. They may carry, touch, grab or attach to something. Usually, we know the shape of those volumes because we know the final function, it might be a sharp edge for a knife or a handle for a tool. Sometimes we must calculate the structural requirements for them to work, in such cases, we must study structural analysis tools such as the ones described in these engineering courses.
  3. The final step is the determination of the linking volumes or connections between all the functional surfaces inside the volume. We must consider the way our printer binds layer by layer to align the main connections to that direction and favour structural integrity ( Sometimes structural anaylises are needed, today there are tools in CAD sofware that can help you do this easily just as the ones used in the example). We must also consider the way our printer lays support material if needed and how it must be removed. The team of researchers proposes the following order to generate such connecting volumes:
    1. Select the main functional surface and design the main connection(s) in alignment to the direction of the layers and the creation of supporting material.
    2. Merge this connection with the functional volume.
    3. Design the rest of the connections among the other functional surfaces (clams, shafts, legs, etc.
    4. If everything joins in one volume, then PRINT.
nvpp_a_679499_o_f0004g
A new global approach to design for additive manufacturing R. Ponche, J.Y. Hascoet, O. Kerbrat & P. Mognol Virtual and Physical Prototyping Vol. 7, Iss. 2, 2012

As we can see, it is important to know our 3D printer and available material very well before designing. We should also mention that implementing the process for the first time is not easy. How can we design something without having an idea of the final result? In order to better understand the process, we will continue with an example.

The case of the bicycle and the Danish Christmas

julecykel
http://www.hareskovcykler.dk

Last year I was very lucky to spend Christmas in Copenhagen. Despite being a small country between Central Europe and Scandinavia, Denmark is an amazingly serious one when you realize their commitment to technological development and implementation. They are one of the leading countries in smartphone diffusion and implementation of moneyless payments (leading Scandinavia and implementing smartphone transactions). Concerning 3D printing, there is a great number of projects and initiatives interested in the implementation of digital manufacturing such as the future fabrication program in the Danish Design Center. Among them, Danish libraries are taking the lead in the introduction of 3D printing through the creation of a nationwide network of FABLABs. I was happy to present the content of this post to the awesome team that collaborates in the Ballerup Makerspace. The Makerspace is a well-equipped workshop underneath the Ballerup library a couple of meters from the train station. It is well equipped with laser cutting, CNC milling machines, and 5 Ultimaker 3D printers completely accessible to the public. It is coordinated by Karen Lykke and a team of 4 people who introduce and orient the public in the use of the Makerspace technology. Here, we made a little exercise that exemplifies the design methodology described above.

Retaking the potential infinite connections that we can design inside the volume of a 3D printer, it is worth trying to mix very strange stuff and prove the global design approach mentioned by Ponche and his team. With this objective in mind, I introduced an ideation method that is traditionally used to create far analogies that inspire teams to create disruptive solutions. This method is commonly known as Forced Fit, Forced Analogies, or Forced Relationships. It works by drawing a concept from a random pool, it can be an image or a word, and use it to inspire weird solutions to our problem. The greater the distance between the random concept and our current problem, the greater opportunities appear for the generation of disruptive solutions. The creative method works by inspiring people to attribute qualities to the problem that they could not think before. Nevertheless, taking in mind that we can connect anything within the limits of our 3D printer, we can dare to combine the different concepts that arise from the Forced Fit method regardless the distance or weirdness of the concepts. For the purpose of this exercise, I decided to introduce two very Danish and very different concepts; bicycles, and æbleskiver, a delicious small pancake-like ball traditionally prepared for Christmas.

Untitled-2
Through brainstorming, we found an opportunity to design an object for drunk-cycling. The design requirements for this object are implicit in the concept itself.

Step 0: Defining the requirements

Using the Forced Fit technique we explored weird pairs of concepts around the bicycle and æbleskiver, we chose concepts in random couples as a starting point for our objects. The example presented here comes from the match between “commuting home” and “gløgg” (a Danish version of mulled wine). As we mentioned before, we imagined an object that could help us drunk-cycling without creating the form itself. The concept of the “Cyggle” as we wrote in a post it, was something that could assist you steering your bike handle safely, therefore the requirements needed were:

  • Have contact with the handle surface to block and slow down its movement (How strong should it be?)
  • Fix the object to the bike frame to create an anchor
  • Hold a beer to continue partying

Step 1: Geometrical analysis

The given requirements gave us an idea of the surfaces that the object will interact with. we know now that we want to block the handle movement, and that the Ultimaker Printer has a printable volume around 20 x 20 x 20 cm. Thus, we compared both volumes and realized that there are 3 functional surfaces that must fit in the printer:

  • The handle brake
  • The frame clamp
  • And the cup-holder

Step 2: Functional volumes

Using the information from our requirements and the available volume, we can define the shapes of our surfaces and volumes if needed. For the initial workshop, we used a combination of paper prototyping and Tinkercad to visualize the fit of the functional surfaces and the printable volume. Next, we proceded to detail the surfaces and volumes that were going to function around the object. For example:

  • Looked for standard 355 ml. beer cans and sized the cupholder to fit it loosely (we want to grab our can and place it back easily!)
  • Measured the size of the handle and frame of an existing bike to scale things properly.
  • Included new features, like teeth in the frame anchor to fix it better.
  • Downloaded a bike model from Thingiverse and fitted everything on.

Step 3: Linking volumes

Untitled-3
From Top-Down – Left-Right: Initial functional surfaces, Functional volumes, Software Optimization, and Optimized form.

As we have already said, we can follow functional volume priorities to design our linking volumes according to our printer layering. However, to make an example of the tools that we can access today, we used Autocad Inventor which in its 2018 version brings a topographic optimization tool that helps us suggest the best linkages for our bodies. Hence, we inserted our functional and printable volumes in the software environment to get a glimpse of the optimized geometry. It is important to mention that this kind of software needs a very detailed version of the requirements, forces, fixed points, materials, volume etc. Form all these requirements, we variated the shape of the printable volume (always fitting into the possible solutions) to play with the computer results. The results created a rib that crossed the volume from the tip of the contact point of the handle to the bottom of the frame clamp. This rib transfers all the force of the movement from one edge to another and therefore should become our main linkage. The cupholder is not as important since it does not experience structural stress. Using this information, we created an optimized form over the software suggestions that would satisfy the structural criteria and have a proper look of a biking product.

Untitled-4
From Top-Down – Left-Right: Initial fin ideas, Fins working sketch, Final fin geometry,  and Curvature analysis that shows how the fins are perpendicular to the main linking volume.

Yet, as you may have noticed, we have a product that lacks the flexibility to turn gradually. It blocks the movement to the extent of the calculated force in the software. In order to make it flexible, we decided to use another feature that makes use of the complexity freedom in 3d printing; thin structures. We divided the rest of the volume into 1.5 mm fins with 1.5 mm spaces between them. These thin structures have been proven to be flexible enough to make simple spring-like structures with common 3D printing plastics like ABS or PET. By dividing this volume, we are facilitating the movement of the handle. Every time the handle pushes, the fins star bending and collapse against the next one. The more you push, the more fins touch, and the more resistance they present. The fins touch each other to the point where all fins are connected and you cannot push anymore. We have aligned the fins to the layering direction of our printer, in this way we guarantee the structural integrity of the linking volume and the lack of supporting material. In this way when printing, the extruder will never interrupt the creation of the layer giving it more strength.

GIF
A simulation that shows how the fins compress and make resistance(I apologize for the moving camera).

As you can see, the final result was not designed before the manufacturing process, in a separated step with assumptions of possible shapes. Instead, the object was created at the same time it was created in the printer environment. This achieved two main objectives; first the connection of two very odd concepts in a real object that could be printed in a Fablab. Second, the inclusion of a greater complexity manifested in more features that include, the cupholder, the frame grip, and the flexible fins. The main point of this post is to demonstrate that even the smallest and most common 3D printer can be used to generate complex proposals. We proof through this example that the 3D printer is a great tool for the exercise of our creative freedom, and a great excuse to experiment and execute innovative ideas. If you have access to a 3D printer, I would like to invite you to try this methodology and experiment.

Untitled-5
Final “cyggle” version

The final “cyggle” version is being printed right now. Once it is ready, an update will be posted with pictures.

I would like to thank the Ballerup Bibliotek Makerspace staff and Karen Lykke for all the time and space provided. 

 

 

Freedom to fabricate … but what?

Today is rare to find someone who does not know what a 3D printer is. The popularity of this technology boomed in 2009 when the first FDM patents expired. Despite being a technology almost 30 years old, it remained a process for industrial prototyping until the expiration of the patents allowed open-sourced manufacturers to create the first desktop 3D printers. Thanks to this release the price of 3D printers fell 55% with desktop printers available for $1,000.00 USD. Such a drastic decrease pumped expectations of journalists and makers who even compared the desktop printer with the Replicator in Star Trek series. In the Sci-Fi series, the Replicator is a machine that rearranges molecular structure in order to synthesize food or other necessary objects for space exploration. Similarly, initial claims suggested that in the future 3D printing would allow everyone to print everything they wanted everywhere they wanted just by pushing a button. Despite not being able to do this yet, much of the popularity of the 3D printer comes from its availability, or what some journalists consider its ability to “democratize manufacturing”.

However, the introduction of 3D printing has gained more traction in industrial instead of desktop applications. Unfortunately, while the growth of the 3D printing industry in 2015 reached $5.1 Billion USD, desktop printers only accounted for an approximate of 25% of the total value of the industry with 278,000 sold units. This means that the real impact of 3D printing can be found in corporations and businesses that currently experiment new applications of prototypes and finished products. For example, according to PWC in 2014, 67% of manufacturers already used 3D printing (2014). Nevertheless, it is important to stress that from those companies that already use printers, just 2.6% use them to fabricate products that cannot be made through traditional methods. Which means that almost all of them are trying to find a way to use it or are just prototyping products that later will be fabricated by other means. Accordingly Michael Stern, from the department of mechanical engineering at MIT, suggests that what we see underlying this apparent boom in the diffusion of 3D printing is a direct substitution fallacy (2015). This means that despite 3D printing can be used to create almost anything just like the Replicator, we are just using it to substitute current manufacturing processes instead of creating something really new.

This raises a huge problem because we already know that additive manufacturing processes (the name of 3D printing in fabrication) cannot substitute traditional manufacturing methods in terms of scale. Comparative analyses have been made by many researchers that confirm that 3D printing is a manufacturing process that must be used to create small and complex batches of products. Economies of scale are a benefit of specializing businesses in just one activity. Making a mold for the production of millions of fidget spinners will always be more effective than 3D printing them. And as we know,  manufacturing a mold requires a huge investment. Unfortunately, when we talk about the “democratization of technology” what we really want is the ability to create freely. However, what we seem to be doing is just doing the same stuff in a very unproductive way which might not help the growth of desktop 3D printing.

 

beast_token
Maybe 3D printing is not as liberating as we thought (Beast Token Fail by http://zheng3.com)

 

Does that mean that the claims of 3D printing potential were all wrong? It is important to say that we have to be always extremely critical with technological hypes suggested by media. The introduction of technology and innovation is always an uncertain matter. Accordingly, predictions and forecasts are more similar to promises that to prophecies. Yet, the properties of the 3D printing process that gave birth to those claims are still true. Therefore I argue that we need to have a better understanding of the affordances that we can access through 3D printing, in order to make a clear image of this democratization of technology and be free to fabricate.

What is really new about 3D printing?

As you might have read somewhere else, 3D printing is a process that allows the creation of freeform bodies. It does so by adding layers of material instead of subtracting it from a big raw piece. This allows it to build precise volumes by placing material where other processes are not able to do the same. Almost all 3D printers use only the material necessary to build the assigned product (some use support material that can be minimized according to the design). This means, that as long as printed products have the same mass, they will have the same cost. This is called “complexity freedom” and is the source of all those extraordinary Replicator claims. Let’s examine what it means.

Complexity is the property of systems that are assembled of many components which have relationships between themselves. The more components and relationships in a system, the more complex the system is. From that perspective, a spoon is not as complex as a pair of scissors. A spoon might have 3-5 components with 2-3 interfaces, whereas a pair of scissors can have 7-10 parts with 5 interfaces which makes it more complex than the spoon. Complexity can also be understood in two categories; functional complexity and manufacturing complexity. Functional complexity is the one that is related to the way the system fulfills its purpose. To fulfill more complex purposes, systems regularly need more components. For instance, the functional complexity of the spoon requires at least one part for food and another for handling. Differently, manufacturing complexity is the one related to the fabrication process. Manufacturing processes and materials have restrictions, which means that  For example, fabricating a wooden spoon allows us to keep the architecture simple, whereas fabricating it with injection molding will force us to include more elements such as ribs and frames that helps the structure of the spoon. In traditional manufacturing, increasing functional complexity usually also increases manufacturing complexity and as a result, fabrication cost. For the same reason, that who owns the more complex part of a product usually gets more revenue from sales. When we say that 3D printing gives us “complexity freedom” is because they break this relationship between functional and manufacturing complexity. This means that when we use a 3D printer, our products can involve a greater number of functional components without increasing the manufacturing cost! Consequently, the cost of production is the same for a cube or for a working clock as long as they have the same mass.

The problem here is that we have always designed our products thinking of a manufacturing process and the restrictions it creates for product complexity. The design process has always been considered a negotiation between the objectives that you have to accomplish with your design, and the possibilities that your resources can afford. There is even a term called Design For Manufacturing and Assembly (DFMA) surged in the 1980’s and 1990’s which looks for the integration of manufacturing knowledge in the design process. So, what shall we do when the process itself has no restrictions? Remi Ponche and his colleagues at the Institut de Recherche en Communications et Cybernetique de Nantes in France suggest that we need to change the way we think of the relationships between our processes and our product. In traditional DFMA we have knowledge of geometries that are possible through the different processes. This pre-defines our available results letting the 3D printer just add a little more value to the design that we have already finished before using the printer. The researchers call this a partial approach because it does not consider the printer completely. What they propose as a global approach is designing with the printer itself. They suggest a process where the volume of the product is developed step by step according to its functional requirements and its position inside the 3D printing volume. Going back to the complexity talk, a process like this helps us exploit the available “complexity freedom”. Allowing us to use the printing volume as an interface that joins all the components that we can insert into it. Therefore, when we claim that the printer can help us create whatever we want, we must correct to say that 3D printing allows us to create complex objects that combine multifunctional components. 

Vive la impression 3D!

Back to the argument for the democratization of technology, we can now say that having a 3D printer at home can give us a tool for the creation of complex combinatory products. But, why would we like to do complex products instead of printing awesome avenger rings? Well, there is a very good reason in what Nobel Prize winner Friedrich Hayek called, the creative powers of a free society. For Hayek, it is very important to accept that despite knowing a lot of things through science, the most important thing we must accept is that as individuals we are pretty ignorant. The world we live in is incredibly complex and it is impossible for everyone to know it all, regardless our academic education. Driving a car is a very good example. For you to drive a car you only need to understand the car controls and the traffic rules. Nevertheless, the car and all the other artefacts, traditions, and institutions around you, have evolved thousands of years by incorporating knowledge of the people that created and perfectioned them. Luckily for you, you only have to take a few driving lessons and pass a test! As this example shows, the progress of our wellbeing is built around the creative solutions that accumulate through time. By people who find a problem and create solutions that later help everyone else. For Hayek, this creative process should be accessible to all. If in the past one person like Leonardo DaVinci, could create so many incredible ideas and solutions just imagine what could happen if we all had the same possibilities that DaVinci had! In a free society, people should have access to creative tools that help them in the creation of new solutions no matter what they are. The 3D printer is one of those tools that have the potential to allow people to combine ideas and create novel solutions for all of us. We just need to learn how to create complex ideas that make use of the capabilities of the 3D printer. 

 

The Additividt Cookbook experiments what 3D printing means beyond technology (http://additivism.org/)

 

Exploiting this complexity is a difficult challenge for all. As we mentioned above, manufacturing complexity has always shaped our thinking. Dealing with this complexity freedom requires that we re-evaluate everything we know about design, manufacturing, and business. Because besides creating solutions for the sake of invention, giving everyone a 3D printer would theoretically allow them to create complex products and profit from them. Economic development initiatives could leverage the impact of 3D printing by giving people access to new methodologies and processes that could help them managing product complexity. Amazing examples can be found in explorations such as the Additivist Cookbook and the Fabricate International Conference Proceedings. Algorithms for topological optimization, structural analysis, parametric modeling, micro-structures, temperature responsive surfaces, and intersections with other manufacturing processes and disciplines are already being developed. In order to completely exploit the possibilities of 3D printing technologies, we need to design systems and interfaces that translate all these developments for everyone in contact with additive manufacturing. Maybe then, the democratization of 3D printing could reach its full potential as a liberating tool. Creators from every point of the society could make use of 3D printers to explore new ways of solving current problems by combining existing artefacts and ideas without the need for capital investment. 

Maybe giving everyone a Replicator to have at home sounds like a great idea. Yet giving away the technology without helping users to understand how it really works sounds more like giving away canned food without giving people can openers. As many experts and journalists claim, 3D printing could start a 4th industrial revolution … as long as with those printers, we also teach people how to design with complexity.

 

What does it really mean to distribute manufacturing?

Last sept 8th the Danish Design Center held a Future Fabrication Summit in Copenhagen’s Carlsberg Byen. The event was part of many very exciting events that happened in the same city since the 5th to the 10th of September 2017 as part of the Techfestival 2017. The summit gathered academics, entrepreneurs, and professionals to discuss the implications of digital fabrication in the creation of self-sustainable cities.

One of the most attractive models for digital manufacturing implementation is distributed manufacturing. Digital manufacturing integrates fabrication processes around a computer. Through computerized control, we can increase the available complexity of the product we fabricate. Robot arms, CNC mills, Laser Cutters, Waterjets, and 3D printers, use digital models to create precise reproductions without molds or human aid. Moreover, information about the product like orders, materials, and designs, can be transmitted as data to remote locations that have the same fabrication equipment and achieve the same results. Distributed manufacturing implements these two advantages in a fabrication scheme where products could be produced in small digital fabrication sites instead of huge centralized facilities. This distributed model could reduce the environmental impact of huge supply chains that move raw materials and inventories around the world. It could also reduce the distance between manufacturers and consumers, creating a more immediate and local model for product fabrication.

By 1983~enwiki at English Wikipedia, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=53526868
By 1983~enwiki at English Wikipedia, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=53526868

Considering its benefits in sustainability, supply chain management, and product design, distributed manufacturing opens many opportunities for new business modeling. Pioneering projects have implemented distributed manufacturing in different modes that range from manufacturing “marketplaces” that connect designers to consumers, to open sourced repositories of models ready for production. Among them, we can find Opendesk, a furniture company that works with a network of manufacturers worldwide in what they call Open-Making. Opendesk designs furniture cut out from standard fiberboards 240×120 cm. Through their e-commerce platform, they allow the customer to select from a catalog focused on workspace furniture, worktables, chairs, bookshelves, and storage units. Images in different materials are available next to an estimated price in local currency. If selected, Opendesk sends a quote request to the associated makers in the network. The price is adjusted to local prices of material and distribution. The units payment is received through Opendesk who distributes all the payments between the designers, the platform, and the makers. The platform has been working since 2014, when they were successfully crowdfunded, adding designs in partnership with studios and including new manufacturers globally.

Opendesk workplace furniture
Opendesk workplace furniture

One of the most interested industries in applying distributed manufacturing on a large scale is the aerospace industry. According to the AirbusVoice Team in Forbes, the demand for spare parts for operating airliners is very slow and irregular. Considering the complexity of an aircraft, the variety of parts that need to be stored in replacement inventory for this uneven demand, implies that inventories today have to be huge. Additionally, available inventory has to be flown to the customer in charge of the replacement causing further delays. Since 2012 Airbus has been experimenting with the use of 3D printers to develop complex parts that can be produced on demand. Spare parts considered suitable for commercial 3D printing such as brackets, and seat lids, have been printed and are currently being tested in operating aircraft. Being able to fabricate on demand for a global industry also would allow Airbus to distribute or even outsource 3D printing of the spare parts. However, current technical requirements for aeronautic regulations have to be addressed in order to continue the development of the business model.

What means to distribute our manufacturing?

Both examples described above show promising opportunities for the creation of new businesses that implement distributed manufacturing. Yet, underlying both examples we can find a problem with product complexity. Complexity in product design and development can be described as a number of elements in a system, and the relationships among them. The more complex, the more parts and relationships they have therefore creating emergent behaviors and novelty. As a result, complex products need more prepared people and complex reliable equipment to be manufactured. Khajavi et al. describe this effect in an analysis of the distributed fabrication of fighter jet parts using additive manufacturing. In their conclusions, the current state of available 3D Printing requires a big investment in equipment and staff to be operational. This means that in order to successfully implement distributed manufacturing today, technology and education should become more accessible. We can see this happening in the examples described above. In the case of Opendesk, the available products are restricted to one method of fabrication and one specific range of materials. The complexity of the product and assemblies is restricted in order to reduce the difficulty in its fabrication. Contrary to Opendesk, aeronautic parts are more complex. Distributing manufacturing operations of an airplane part requires matching very specific technical requirements that make flying safe. Thus, if we want to distribute the fabrication of airliners’ spare parts we must upgrade each of the manufacturing sites to comply with all the technical standards. Then, we can say that distributed manufacturing needs of product complexity management to be successful. 

Managing product complexity is in itself an important business matter. According to the theory of incomplete contracts, firms try to include inside themselves the transactions that they can’t control through a contract with other entities. A complex product, such as a new technology, or an ad-hoc solution cannot be completely split and described between two partners and as such, it must remain in control of the company that has invested in its creation. Again, this shows the difference between the two examples above. While making furniture does not deal with state of the art technology, Opendesk is able to share the information with their suppliers. Every partner can clearly understand where their responsibility starts and ends as much as how much they are being paid. In the airplane spare parts example, the complexity of the parts would make a huge deal if distributed because processes such as titanium 3D printing have many variables such as controlled atmospheres, and low-dimensional tolerances. Hence, the communication between two parties (inside a company or between two companies) would be very complicated to manage. A company just like Airbus would have a hard time sharing information with distributed partners. Subsequently, the problem of distributed manufacturing is not only a technological problem, is a management one too. Unfortunately, this management problem is poorly addressed when we discuss digital fabrication with the general public, and when we offer it as an alternative to existing business models.

 

LEGO_42055_WEB_PRI_1488
More complexity brings more components and relationships between them. (Lego 42055 BUCKET WHEEL EXCAVATOR)

 

Back in the future fabrication summit in Copenhagen, Nat Hunter Strategic Director of the Machines Room stressed the role of the designer as an actor of this new production model. Many other authors highlight this as an effect of digital technologies that combine the roles of designer and the manufacturer. We can see how digital technologies bring together activities that before were split into planning and building thanks to advances in simulation and the precision and speed of these new fabrication methods. We have shown above how product complexity management is one of these activities in planning that needs to be addressed if we are going to implement digital manufacturing in a distributed model successfully. Maybe it means that managers need to bring designers close or even become designers in order to create new ways of going around the problem of product complexity in distributed models.

Questions for the future

Considering managers as designers would open very interesting questions that can push the implementation of distributed manufacturing forward. What is the role of design in enterprise management software? How can we manifest product complexity in ways that facilitate talking about it and sharing? How do roles inside a digital manufacturing enterprise should evolve to accommodate distributed manufacturing? Which is the place of product design inside management and entrepreneurship? However, the main question that we have to answer is the question of innovation. Innovation is the main differentiator of new businesses, it makes space for them to survive and create new market opportunities. As mentioned above, novelty is commonly related to product complexity. Hence, novel businesses are usually complex as a result of their innovative solution. Does this mean that there can’t be new product development through distributed manufacturing? Are distributed manufacturing models only possible if we manufacture simple products? More work is needed in order to answer these questions. Luckily, projects such as the FabCity prototype in Poblenou, Barcelona and the Maker Mile in London will throw some clues that will help us evaluate if distributed manufacturing is a viable alternative for new business making.