A team from ETH Zurich research institute has transformed a 12-ton Menzi Muck M545 walking excavator into a robot named HEAP (Hydraulic Excavator for an Autonomous Purpose). HEAP was equipped with GNSS global positioning, an IMU (inertial measurement unit), a control module, and LiDAR sensors for precise navigation and manipulation. HEAP's task involved building a 6-meter high, 65-meter long dry-stone wall using randomly shaped boulders. The robot started by scanning the construction site, creating a 3D map, and then analyzing the boulders' weight, center of gravity, and three-dimensional shape using machine vision. It determined the optimal placement for each boulder to construct the wall without any mortar, placing approximately 20 to 30 boulders per building session, allowing for the use of locally sourced materials to save energy.

‍

Building a dry-stone wall with randomly shaped boulders is a physically and mentally demanding task, making it a prime candidate for robotic assistance. The HEAP excavator, a modified Menzi Muck M545, successfully completed this challenging task using advanced technology such as GNSS, LiDAR sensors, and machine vision. By scanning the construction site and strategically placing boulders, HEAP constructed a stable dry-stone wall, reducing the need for energy-intensive transportation of building materials. This innovative project demonstrates the potential for robots to take on complex construction tasks in the future.