Amsterdam: Redesign and Machine Learning Will See MX3D Complete 3D Printed Bridge Next Year

[Image: Joris Laarman Lab]

MX3D is famous for a unique 3D printing process that not only extrudes filament straight into thin air but is industrial strength too. With no need for support structures and the ability to 3D print with metal, MX3D developed Wire and Arc Additive Manufacturing, or WAAM. And while the uses for 3D printing and the innovations that can be created already seem infinite, MX3D allows even further freedom in the creation of large-scale objects.

We first learned about MX3D as they embarked on a journey in Amsterdam to consider making a 3D printed pedestrian bridge, and then worked with Dutch students fabricating a 3D printed metal bicycle. Fast forward two years later, and now the group is back to completing the bridge that they began work on with Joris Laarman in 2015 (with Autodesk as chief investor). Progress should be rolling along more quickly though as the team expects the bridge to be finished in 2018.

[Photo: Olivier de Gruijter]

Since the inception of the idea for the bridge in 2015, the teams have overcome numerous obstacles. They also realize that to continue progressing they will have to improve on current software, as well as robotics; in fact, they foresee robots imbued with the ability to learn and improve (especially at 3D printing) as a requirement for future applications that the bridge actually serves as a proof of concept for.

“We’re now making huge steps in the volume of objects that can be printed. That’s going to create a significant leap in adoption,” says Gijs van der Velden, who runs MX3D.

Laarman originally envisioned the bridge as a system of struts networking out in an open, almost crystalline design. And although the initial plan was to install the 3D printed bridge over Amsterdam’s famed Red Light district, changes were required because the first design would have placed undue stress on the canal walls.

Refinements to the bridge mean that it will be a more traditional pedestrian structure, though enhanced by complex bends and turns—accents which would not have been possible without 3D printing. Robots are expected to do a great deal of the construction work also, thanks to innovative software. 3D printing massive pieces though presents a challenge, especially considering the nuances of making a structure with steel.

Because of the intricacies involved, the machines (the robots) must be able to learn. Right now, sensors alert the robots regarding how much current is used to heat the metal to the appropriate temperatures. They are also able to understand how hot the metal with get, and where to apply welds.

MX3D will take machine learning one step further as they design algorithms allowing robots to understand which welds could be problematic and then work to fix the issues—or even better, avoid them ahead of time.

“When you’re making the file for printing, the big issues will be resolved,” explains van der Velden. “When you’re actually printing, the machine will recognize a problem and create a solution on the fly.”

Van der Velden does point out, however, that this type of technology will probably only be applied to a sliver of industrial construction. Their niche will be in complex constructions, probably amounting to five percent of projects overall. Oil rigs are a good example where the technology would be helpful, with robots doing the bulk of the work under human supervision. With 3D printing, parts will be more lightweight, affordable, and durable—as well as high-performance.