unit-code
CO-SPACE is a project to provide housing for members of cooperatives, which operates through a hybrid platform with the engagement of local organisations. As a living architectural system, CO-SPACE centres around the creation of flexible spaces facilitated by a bottom-up system that enables self-organisation, configuration, and negotiation. The system is based on wave function collapse and spatial planner algorithm.
Furthermore, leveraging artificial intelligence and machine learning, modular collaborative robots will construct and modify the various spaces according to the requirements of their occupants. The project aims to achieve coexistence between humans and robots. In the CO-SPACE ant-cave-like building space, humans and modular robots resembling a colony of ants and can live together.
CO-SPACE is a cooperative housing project with a modular construction robotic system, space-generating algorithm system, and interaction platform system.
The construction module consists of a uniform prefabricated panel, a central connector for robotic link, and an edge connector for joining the modules.
The project employs three module geometries, enabling diverse assembly methods. A small-scale wave function collapse algorithm was used for automatic assembly and module number count.
The robots are connected to each other by grippers. Three types of robots have different gripper setups and thus a variety of different combinations are possible. These further influence the way robots co-operate during construction activities.
Three types of robots work together in the construction process. Robots work with robots and with modules to perform free movement on horizontal floors and vertical walls, as well as translation and handling of modules.
After determining their form, the team experimented with the physical robots for activities such as module transport, robot lifting, and obstacle avoidance during collaboration.
Setting the user agent colour and placing it into the algorithm enables the automatic generation and negotiation of each spatial region. The algorithm is also capable of allocating space to specific regions during user migration.
The project designs unit space tiles, classified according to the privacy, extensibility, and regularity of various space types. A large-scale wave function algorithm computes them and generates the final spatial form.
Using a uniform spatial grid for the spatial planner and wave function collapse algorithms allows simultaneous computation. The spatial planner generates regions fed into the wave function collapse to create specific spaces in corresponding areas.
CO-SPACE offers co-op housing. Rent payments are split between CO-SPACE and the co-op operation. Rent factors in room volume and robotic activity. Shared spaces allocate rent based on members' incomes.
The Co-Space platform collects user data to build communities and generates cooperative areas, shared areas, and private areas in suitable locations. Detailed spaces are generated according to the needs of different types of spaces.