
3DLOOP
CAMERA-ROBOTIC SENSING FOR 3DCP
Project ID

PROJECT BY
PARTNERS
YEAR
SITE
TYPE
TEAM
ROLE
SDU.CREATE, Led by Assoc. Prof. Dr. Roberto Naboni
Hyperion Robotics, Weber Saint Gobain
2022-2023
Odense, South Denmark
Research Project
Luca Breseghello, Roberto Naboni, Daniele Florenzano
Researcher, Computational Designer
Project Description
Construction Robotics is a rapidly upcoming area of development in the field of automation. The building industry suffered historically from insufficient automation levels, which are explained by the difficulties related to automating material-intensive construction processes, the scale and complexity of the construction environment, and the level of customization needed for building production.
Extrusion-based 3D Concrete Printing (3DCP) is one of the leading development areas in architectural and construction robotics at present. Despite its great potential in revolutionizing the way we build sustainable architecture and providing a safer construction site, the robotic 3DCP process is still characterized by a limited automation level. This is because the high degree of complexity in handling cementitious materials still demands human interventions to manage changes in operational conditions, e.g. temperature and humidity.
This project aims at rendering robot-based 3DCP autonomous and robust by 1) monitoring key aspects that impact the 3DCP process and 2) elaborating a concept for a robotic system capable of adjusting in real-time aspects such as the extrusion rate as well as quality and dimensional monitoring.

Monitoring Scenarios
The experimental activity developed has the goal of validating a monitoring approach for 3DCP in various geometrical scenarios. Different cases of variable complexity have been devised, although the analysis has been performed only on Case 1.

Baseline TEST Design

PRINTING ACTIVITY
The printing experiment consists in the production of a sample of 8 layer of the baseline design geometry to collect information and test material defects.
SCANNING ROUTINE
After each layer, a scanning routine is performed to collect a high resolution point cloud at a each specific scanning point.
DATA COLLECTION AND ANALYSIS
After collecting the data, the point clouds are post processed and analysed to collect information of the layer geometrical properties (i.e. height, width and layer deviation).
_gif.gif)
