CoCoBot: Collaborative Construction site roBot for assistive logistic tasks
This project focuses on exploring the possibilities for automation in the construction industry using mobile robots for assistive transport of tools and materials.
Background
This project focuses on exploring the possibilities for automation in the construction industry using mobile robots for assistive transport of tools and materials.
Construction relies on human operated tools and machinery thus being one of the most underutilized sectors in terms of robotic automation. While other sectors of industry are largely benefiting from the merits of robotic automation, construction applications are restricted due to challenging operating environments with ever changing landscapes, harsh weather conditions and safety requirements. Thus, providing solutions with current technologies in mobile robotics can provide future opportunities for optimizations in the construction industry.
The project will be exploring the use-case of carrying long beams of material, e.g., plastic, wood, or metal, that will be loaded on an existing mobile platform rated for outdoor use.
By pulling on one end of the beams, the worker will guide the robot into a desired location in the construction site. The robot will interpret the forces and torques applied to the beam as steering and navigation commands. Using sensor fusion and state-of-the-art algorithms, the robot will be endowed with advanced navigation capabilities. This will allow the robot to adjust its path dynamically in order to avoid obstacles, humans and terrain deformations. When the load has reached the desired location, the worker will release the beams, causing the robot to stop due to lack of user input. The worker can then set the robot on a "halt" state in order to perform the offloading process of the carried material.
Purpose / Vision
There are lots of logistic tasks to be carried out in a construction site, currently being handled by a single person in case of small loads, or two or more people for bigger items. This Jack of automation results in hazards for the workers, as well as, a costly and time consuming construction process. Therefore, a building constructor and building sub contractors (e.g. carpenters) will help elucidate requirements and current needs in this field as well as plan possible future use-cases. Using provided proving grounds, the project is ensured to address real-world problems and be tested in relevant environments in direct collaboration between all project partners - where several SMEs are expected to join as technology providers.
Enabling a mobile robot to navigate a construction site and help with these logistic tasks is the first step on the road to solving more complex tasks using robotics in construction. Thus, testing how existing mobile robots can be adapted with additional sensors and with suitable perception algorithms to perform a variety of logistics tasks in construction sites.
Since construction sites are changing rapidly and often contain many different types of uneven terrain, obstacles, or even weather conditions, standard mobile robot localization and navigation methods are not suitable for this type of environment. Therefore, an appropriate mobile platform will require multiple sensing modalities operating in a complimentary manner. This project aspires to identify the minimally required set of modalities and how to fuse them in the most robust and effective manner. The processing of sensor data and extraction of high-level information about the surrounding environment, i.e., position of obstacles and humans, are vital for this application. These processes will allow research partners to develop and integrate technologies and explore current limitations.
Employing human workers in a collaborative task with the robots will increase the acceptance of robots as viable tools on a construction site. Therefore, an appropriate human-robot interaction method must be natural and close to the collaborative task with other workers. To this end, the research partners are going to define a robot control interface based on forces and torques applied on the carried materials and identify possible additions or/and revisions to the human-robot interaction process. Possible future exploitations will include the robot guiding the human co-worker while avoiding obstacles and the execution of the same collaborative task performed solely by multiple robots.
Expected results
The project will allow the end-users to obtain a real-world indication of the future potential for robotic technologies as an integral part of the building construction process, and the technology providers will gain significant experience in the use of their products in new domain uses. Additionally, the results of the project will be shared through news publications and videos, which will be used as steppingstones for future collaborative research proposals. From their participation all partners will identify future challenges and engage common efforts to resolve them in future research projects.
Project participants
Teknologisk Institut Robotteknologi
Funding
RoboCluster has funded 350.000 DKK for the project.
The project will run from January 2020 – September 2020
Contact
Do you want to know more about the project?
TI (lead): Tsampikos Kounalakis tsko@teknologisk.dk
DTU: Lazaros Nalpantidis lanalpa@elektro.dtu.dk
Want to learn more about this project?
For more information about this project, please contact Project Manager Ole.
Co-project Manager, Side Events