ShotcreteDepth: A Bi-modal Dataset for Robust Robotic Depth Perception in Shotcrete Construction Environments
2026-06-22 • Robotics
Robotics
AI summaryⓘ
The authors created ShotcreteDepth, a dataset with images and 3D scans from construction sites where shotcreting is happening. Their data shows tough conditions like dirty air and low light, which make it hard for machines to see properly. They also made a simple tool to help label the 3D data faster. The dataset includes over 11,000 data points, some with detailed labels for testing. It helps researchers improve how computers understand depth in challenging real-world industrial settings.
ShotcretingStereo RGB imageryLiDAR point cloudsDepth estimationDepth completionStereo matchingAutonomous perceptionAnnotation toolTemporal synchronizationConstruction environment
Authors
Jakub Gregorek, Lars Arnold Dethlefsen, Patrick Schmidt, Mads Essenbæk, Jonas Flink Bentzen, Lazaros Nalpantidis
Abstract
We introduce ShotcreteDepth, a bi-modal dataset from the construction domain that captures both an active shotcreting process and general construction environments. The dataset comprises stereo RGB imagery and LiDAR point clouds acquired under harsh real-world conditions, including high turbidity and poor illumination. Such conditions adversely affect sensor measurements, leading to incomplete and noisy observations that pose significant challenges for perception systems in autonomous applications. Alongside the dataset, we release a lightweight annotation tool designed for time-efficient labeling of LiDAR point clouds. ShotcreteDepth consists of 11,252 temporally synchronized data samples, of which 220 are annotated for evaluation purposes. The dataset supports research in stereo matching, depth completion, and depth estimation under conditions that closely reflect the operational complexities found in industrial settings. Project repository: https://github.com/dtu-pas/shotcrete-depth