Collaborative Construction

CoConstruct (Collaborative Construction), coordinated by Prof. Dr. Christoph Gehlen, explores novel collaboration concepts between humans and machines to improve productivity, resource efficiency, and increase quality in the Architecture, Engineering, and Construction (AEC) domain. A team of researchers from various disciplines, including civil engineering, mechanical engineering, material science, computer science, and architecture, will explore innovative ideas for automating design, planning, and construction using a human-in-the-loop approach. This involves combining human skills - cognitive and sensory abilities - with Machine Learning and robotics techniques.

In the context of design and planning, the CoConstruct approach requires developing novel computational design methods which allow integrating quantitative metrics and the analytic power of computation (such as structural integrity, energy performance, life-cycle assessment, acoustics, and cost estimation) with human synthetic thinking and creativity. In the context of fabrication and construction, the CoConstruct approach will require to leverage advanced perception capabilities and spatial AI to investigate how several robotic units can cooperate and support each other or take on individual assembly tasks in direct and indirect cooperation with humans, and thus allow for novel robotic construction processes to be carried out in turn-taking physical action by human-robot teams. Based on a case-study methodology, the research will develop novel methods which will be experimentally explored and validated based on real-world scenarios.

Our Team

Doctoral theses

  • Lidia Atanasova (Building (with) Human-Robot Teams: A Framework for Human-Robot Cooperative Assembly Processes)
  • Lazlo Bleker (Machine learning-assisted computational structural design)
  • Daoyi Gao (3D Scene Understanding from Retrieval and Object-centric Perspective)
  • Mohammed Olabi (Automated joining technologies for additively manufactured concrete structures)
  • Pia Stanzel-Deffner (Agile Methoden in der Immobilienentwicklung)