3D printing, or additive manufacturing (AM) as it’s more appropriately termed by the industry, has been around for at least 30 years since the commercialization of stereolithography in 1987. It holds the potential of mass customization, where products are uniquely designed and produced specific to their final use or user, at scale. But despite the long history of AM and the quantifiable benefits, it’s still predominantly used for prototyping purposes.
There are a few examples of AM moving from prototyping to production – for true mass customization the technology has a strong hold in dental and is starting to take hold in footwear. Both are applications with high volume worldwide demand, previously met by mass-produced products that fit no-one well or expensive custom products available only to the wealthy. Another example of AM in production is for small series high performance engine parts, particularly in metals, for aerospace and automotive. But the idea of an additive factory that produces engines one day and medical imaging devices the next day is still far from reality.
The reason why isn’t a single factor, but rather a series of issues that have compounded and kept this decades-old technology from revolutionizing manufacturing. In this article, we’ll discuss the five factors that we see as most important to the adoption of industrial additive manufacturing, and what needs to change in order for it to realize its potential.
Every challenge on this list of course leads back to cost, at the end of the day AM won’t be broadly adopted and used at production scale unless the ROI exceeds other available production methods.
But the high cost of the machine itself is a primary issue. Machine cost can account for between 60 to 80 percent of total AM expenses. For production-quality metal AM, a “low cost” machine can be had for around $200,000, but depending on the manufacturer and details such as build volume and number of lasers, the cost can increase to upwards of $2 million per machine.
AM machine manufacturers currently spend significant money on research and development, and these expenses are reflected in the cost of the machines. While R&D definitely helps move the industry forward, as the technology becomes more mature R&D costs should also decrease, resulting in lower machine costs.
With current technology, AM can easily cost between $50 to $100 per hour. The average build can take a few hours of machine time, although it can range from minutes to months depending on the size. Improvements in both AM software and hardware that increase output speed – and therefore capacity – will help reduce the capital cost of the machines.
Competition will also help drive down the cost of the equipment. New entrants to the market take advantage of expiring patents and new technologies to release products at lower price points, putting pressure on incumbent pricing and business models.
2. Repeatable Quality
To be viable for production, part quality must be predictable and repeatable during additive manufacturing. This is critical in order to increase production output and reduce waste, which of course reduces cost. Whether we’re making a single item or multiples, the precision and quality must be directed and controlled across the process.
Factors that influence part quality and repeatability include:
- Temperature and humidity in the room where the machine is housed
- Consistent material quality from batch to batch
- The need to refill material during the process, causing a change in part properties
- Different placement on the build platform, since the laser is most accurate in the center of the platform
- For metals, the different combination of parts on the build platform resulting in different temperatures and thermal stresses
- Post-processing required after the build is finished can introduce variables and opportunities for error.
Post-processing not only increases the opportunity for error but is also a dependent step that can be streamlined. Automating tedious tasks such as support removal and polishing could not only reduce potential for error, but also reduce time and cost.
Today, all of these tasks listed above are technically possible to overcome but are currently solved by a combination of manual intervention, a heroic effort on the part of the AM engineers (see below), or over-compensation for potential shortfalls, all resulting in higher costs.
3. Availability of Trained Engineers
According to Deloitte, 9 of 10 additive manufacturers are struggling to find the skilled workers they need. To some degree, you could say AM is a victim of its own success. According to Wohlers Associates, the average compound annual growth rate (CAGR) of metal 3D printing has been over 30 percent over the last ten years, and an incredible 80 percent growth in 2017. As the industry continues to boom, it’s likely that the worker shortage will continue to pose problems.
The shortage is also unsurprising due to another factor: for a technology that’s been around 30 years, AM still requires a great deal of specialized knowledge and expertise, from how to design for AM; to calibrating, operating and maintaining AM machines; to understanding material handling and performance, to integrating AM into a production facility. Poorly trained or inexperienced engineers make inefficient decisions and avoidable mistakes that increase the average machine time per part and therefore the cost per part.
Training AM engineers for these specialized skills is an expensive and time-consuming investment. Respectable training programs for metal 3D printing involve at least 6 months of trial and error, which is costly not just for the human time but also machine time and material use.
While training and hiring more engineers is a priority, the flip side of this coin is that both the hardware and the software technology simply must become smarter on its own and be more intuitive to use so that the high level of expert intervention isn’t required – just as we expect from other mature technologies.
4. End-to-End Solutions
Hardware and software solutions for industrial applications of AM are making progress, but as already stated they are too difficult to use, requiring expert technicians to operate. The other issue is that most have developed to focus on addressing specific fields or applications, such as dental or aerospace. In order to be viable as production applications, these solutions must develop both breadth and depth. They have to be powerful and flexible enough to be used for the wide variety of applications that our “future factory” is intended to support.
As it stands today, companies need to use multiple software packages in order to go from design to production:
- A computer-aided design (CAD) program is used to make the model
- Though some CAD tools are starting to include basic build preparation features, most professionals use a separate computer-aided manufacturing (CAM) software tool to prepare the design for printing and output a machine-readable file.
- Next, planning software is used to schedule the job on the machine
- Finally, technicians adjust the settings on the machine itself to optimize for a successful print
Each of these tools has its own level of complexity and requires specific expertise, presenting a barrier to adoption. Also, each handoff between steps takes time, and creates opportunities for introduction of deviation or errors.
Some AM hardware companies have launched initiatives to build their own software tools, and many CAD companies are moving toward developing AM software or features. But these remain specialized solutions, focused on discrete applications or machines. None yet represents a comprehensive and seamless end-to-end solution that can help speed AM adoption.
5. Integration with Other Processes
Finally, though often the stories of a utopian additive manufacturing ‘factory of the future’ portray ‘push-button manufacturing’ and products coming out of the machine ready to ship, the reality is that AM will always be one process in a series required to make something.
Even with wholly additively-produced products, finishing is needed after the build is complete, such as cleaning, removal of supports, and polishing. Understanding what kind of post-processing is needed for which builds currently requires a technician with mechanical expertise, which adds complexity into the integration process.
Also, these interdependent processes must all be planned and scheduled, including the development of any required tooling, safety procedures for handling or disposing of materials, and of course the machine time, setup, cleaning, and technician’s time. Any problems encountered at a step mean that subsequent steps experience delays or must be rescheduled. With information generally siloed at each specific step, disconnected downstream processes don’t get informed by learnings or adjustments made earlier, adding to the cost and complexity.
In order to become production-ready and see wide adoption, factory processes need a way to be seamlessly integrated at every level, informing and updating each other to streamline and ultimately reduce the cost of production.
The Road Ahead
While progress is being made in industrializing AM, a lot of work still remains to be done. Above all, cost needs to come down so that the ROI of mass customization exceeds that of mass production, at least for certain appropriate goods and markets. As the technology advances, if we focus on solving for the five factors above, we can expect to see greater industrial adoption of AM in the coming years.
 “Costs and Cost Effectiveness of Additive Manufacturing.” NIST. December 2014
 Vazquez, Passaretti and Valenzuela. “3D opportunity for the talent gap.” Deloitte Insights. March 24, 2016
 “Wohlers Report 2018” Wohlers Associates, Inc. March 27, 2018