Dutch robotics company MX3D has completed the 3D printing of a steel bridge, which will be installed across a canal in Amsterdamnext year.
Designer Joris Laarman worked with the robotic manufacturing technology start-up to build the 12-metre-long pedestrian bridge, which is being previewed at Dutch Design Week between 20 and 28 October.
First proposed in 2015, the structure has been constructed by robots from layers of molten steel. Its span was completed earlier this year. Now its deck has been built, marking the final stage in the construction process.
Six-axis robots built the six-metre-wide structure from layers of molten steel, which involved programming robotic arms to control large-scale welding machines.
The ambitious project has gone through several iterations in its development to strengthen the structure against potential boats collisions and to adhere to the local council’s regulations.
“The initial design changed significantly due to the engineering concerns,” said MX3D. “The lightweight tree-like structure is built on four main bearers. The concern was that if one of these bearers gets hit by a boat, the structure could critically fail.”
“Therefore a totally new concept was created that took into account everything we learned until then about the material, the local specifications and regulations,” explained the company.
MX3D originally meant for the robots to print the bridge on site, beginning with the four structural bearings and working inwards, but the plan had to be abandoned because of budget restraints and as it was “just too dangerous”.
“We have set a vision in which robots would be able to autonomously 3D print infrastructural interventions in our built environment,” said a spokesperson.
“Over the last three years, we have been able to create the technology required for this plan but we did not get permission to do it by the city government.”
“Our timeline and budget also made it hard to realise this vision so we focused primarily on 3D-printing a fully functional pedestrian bridge, including a sensor network,” they continued.