Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Surface Quality Assessment in Mobile Robotic Construction
Construction Robotics is an interdisciplinary field at the intersection of architecture, civil engineering and robotics. In this project, you will work together with experts from all fields using a robot that performs high-quality surface finishing.
Keywords: robotics, computer vision, high-precision sensing, digital fabrication
Construction Robotics is an interdisciplinary field at the intersection of architecture, civil engineering and robotics. In this project, you will work together with experts from all fields using a robot that performs high-quality surface finishing.
Every building process is unique in terms of the materials used and their behavior as well as the range of acceptable tolerances of the elements it produces. The methods that facilitate an additive robotic construction workflow need to be tailored to sense various features of the materials used and the resulting surface qualities. RPS (robotic plaster spraying) process explores the material informed design process of bespoke surfaces, combining an off-the-shelf cementitious plaster mix with an adaptive, on-site fabrication process. In this process, a 6-DoF robotic arm is used to spray plaster onto a target surface, which can be used for data collection through physical testing. This data collection involves 3D scanning of the target surface to obtain information on the volumetric or textural formation in relation to the fabrication parameters. This workflow allows for retrieving surface geometry to enable adaptive fabrication and to collect data to visualize the effect of the process parameters on the resulting surface geometry / pattern through the use of both linear and non-linear prediction models. These models can then be implemented in an intuitive digital (design and visualization) tool that can support the design process with a malleable material such as plaster.
In this project, the goal is to collect data with systematic tests (small-scale samples with a stationary setup or from larger prototypes with the mobile robot). Like this the addition of surface quality to the design and visualization tool that employs a non-linear regression model will be explored. The workflow consists of retrieving surface quality (intensity maps, 3D point clouds) using i.e. a time-of-flight-camera mounted on the robot end-effector, registration and stitching of multiple scans, processing based on reflectance, feature extraction and relating features to the parameters of the robotic spraying process.
Construction Robotics is an interdisciplinary field at the intersection of architecture, civil engineering and robotics. In this project, you will work together with experts from all fields using a robot that performs high-quality surface finishing.
Every building process is unique in terms of the materials used and their behavior as well as the range of acceptable tolerances of the elements it produces. The methods that facilitate an additive robotic construction workflow need to be tailored to sense various features of the materials used and the resulting surface qualities. RPS (robotic plaster spraying) process explores the material informed design process of bespoke surfaces, combining an off-the-shelf cementitious plaster mix with an adaptive, on-site fabrication process. In this process, a 6-DoF robotic arm is used to spray plaster onto a target surface, which can be used for data collection through physical testing. This data collection involves 3D scanning of the target surface to obtain information on the volumetric or textural formation in relation to the fabrication parameters. This workflow allows for retrieving surface geometry to enable adaptive fabrication and to collect data to visualize the effect of the process parameters on the resulting surface geometry / pattern through the use of both linear and non-linear prediction models. These models can then be implemented in an intuitive digital (design and visualization) tool that can support the design process with a malleable material such as plaster.
In this project, the goal is to collect data with systematic tests (small-scale samples with a stationary setup or from larger prototypes with the mobile robot). Like this the addition of surface quality to the design and visualization tool that employs a non-linear regression model will be explored. The workflow consists of retrieving surface quality (intensity maps, 3D point clouds) using i.e. a time-of-flight-camera mounted on the robot end-effector, registration and stitching of multiple scans, processing based on reflectance, feature extraction and relating features to the parameters of the robotic spraying process.
Research on the state-of-the-art.
Develop methods for registration and stitching and to generate features that represent resulting surface textures and qualities.
Experiment with different sensor setups.
Collect data sets for training the non-linear regression model.
Assessment of the proposed data collection method by on-site testing with the stationary and the mobile robotic setup.
Research on the state-of-the-art. Develop methods for registration and stitching and to generate features that represent resulting surface textures and qualities. Experiment with different sensor setups. Collect data sets for training the non-linear regression model. Assessment of the proposed data collection method by on-site testing with the stationary and the mobile robotic setup.
Highly motivated and independent student.
Strong interest in 3D Sensing, Measurement Systems, Construction Processes, Hardware Experiments.
Programming skills in C++ or Python are mandatory.
Experience with ROS, COMPAS, cgal, Open3D, etc. are a plus.
Highly motivated and independent student. Strong interest in 3D Sensing, Measurement Systems, Construction Processes, Hardware Experiments. Programming skills in C++ or Python are mandatory. Experience with ROS, COMPAS, cgal, Open3D, etc. are a plus.
If you are interested in this project, please send your transcript and CV to:
Hermann Blum hermann.blum@mavt.ethz.ch
Valens Frangez valens.frangez@geod.baug.ethz.ch
Selen Ercan Jenny ercan@arch.ethz.ch
If you are interested in this project, please send your transcript and CV to: Hermann Blum hermann.blum@mavt.ethz.ch Valens Frangez valens.frangez@geod.baug.ethz.ch Selen Ercan Jenny ercan@arch.ethz.ch