United Arab Emirates-based design practice MEAN* (Middle East Architecture Network) has proposed a sculptural 3D-printed pavilion for a prominent traffic roundabout to welcome visitors to the upcoming 2020 Expo in Dubai. Designed as a “spatial forest,” the interactive installation comprises a series of palm tree-like concrete elements and branching LEDs. The “Expo 2020 Landmark” proposal is also powered with solar energy and can be programmed to light up at night with a variety of lighting modes. 

Continue reading below
Our Featured Videos
round and tear-drop shaped columns topped with solar panels

Towering at a height of over 26 feet, the domed Expo 2020 Landmark was inspired by the Expo 2020 logo and UAE’s iconic palm trees. As a symbol for innovative construction, the installation would be built from 3D-printed shell components that can be cast on-site with Ultra High Performance Concrete, a material selected for its durability and resilience to Dubai’s harsh desert climate.

Related: Energy-producing pavilion proposal for Expo 2020 mimics Brazil’s biomes

several white, 3D-printed columns
rendering of person walking toward pavilion made of 3D-printed columns

“Robotically 3D-printed concrete construction has been lauded for saving on material waste by reducing the amount of formwork involved in the process of casting, as well as providing a cleaner construction site, all while allowing for a higher degree of complexity in design,” the architects said in a project statement. “We believe that Expo 2020 would be a fantastic platform to showcase the possibilities of this emerging construction technology to the world.”

curved, white columns
white pavilion glowing at night

The Expo 2020 Landmark can be enjoyed by motorists traveling in the roundabout as well as pedestrians, who would be invited to enter the pavilion and explore the spaces between the 3D-printed, palm tree-inspired elements. Solar panels installed on the structure would be strategically tilted for maximum solar exposure and to deter sand buildup.


Images via MEAN*

several thin white columns