Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Perception and manipulation for wooden shingle envelopes construction using the quadrupedal manipulator ALMA
The goal of this project is to develop an integrated framework for the detection and pick-and-place of irregular wooden shingles with rough surfaces using the quadrupedal robot ALMA for automating the assembly process of wooden shingle roofs.
Keywords: Mobile manipulation, perception, motion planning, construction robotics
Legged manipulators are becoming more reliable to be used in the construction industry to automate architectural building processes. Autonomous construction in dangerous and hazardous environments can provide significant advantages in terms of human safety and operational opportunities. However, the robotic assembly of raw building materials presents several challenges. For instance, irregular building materials often come in diverse shapes, sizes, and configurations. This variability makes it challenging for robots to handle and manipulate them accurately. Additionally, each piece may require custom analysis and adaptation by the robot. Thus, the assembly process relies on accurate perception, localization, planning, and control technologies.
In this project, we aim to use the quadrupedal robot ALMA for enhancing the perception and grasping of non-standard and unprocessed wooden pieces called shingles. Some algorithms have already been developed for the perception and object detection of shingles which could be reused in this project.
Legged manipulators are becoming more reliable to be used in the construction industry to automate architectural building processes. Autonomous construction in dangerous and hazardous environments can provide significant advantages in terms of human safety and operational opportunities. However, the robotic assembly of raw building materials presents several challenges. For instance, irregular building materials often come in diverse shapes, sizes, and configurations. This variability makes it challenging for robots to handle and manipulate them accurately. Additionally, each piece may require custom analysis and adaptation by the robot. Thus, the assembly process relies on accurate perception, localization, planning, and control technologies. In this project, we aim to use the quadrupedal robot ALMA for enhancing the perception and grasping of non-standard and unprocessed wooden pieces called shingles. Some algorithms have already been developed for the perception and object detection of shingles which could be reused in this project.
- Literature review on vision-guided robotic assembly of non-standard objects
- Perception for detecting shingles
- Motion planning
- Literature review on vision-guided robotic assembly of non-standard objects - Perception for detecting shingles - Motion planning
- Good knowledge of C++ and Python
- Experience with vision and perception
- Experience with ROS
- Good knowledge of C++ and Python - Experience with vision and perception - Experience with ROS
Send an email with your CV and transcripts to Jia-Ruei Chiu at **jichiu@ethz.ch** and also CC Farzaneh Eskandari at **farzaneh.eskandari@epfl.ch**
Send an email with your CV and transcripts to Jia-Ruei Chiu at **jichiu@ethz.ch** and also CC Farzaneh Eskandari at **farzaneh.eskandari@epfl.ch**