Haku

Machine learning algorithm for automatic routing of underfloor heating circuits

The machine learning based routing engine enables automatic design of underfloor heating circuits.

  • Generate various circuit options
  • Test routing alternatives by adjusting parameters
  • Select a suitable option and install the circuit to the room

The Auto-Routing Underfloor Heating function employs a machine learning algorithm to produce optimized underfloor heating layout suggestions for users. The function considers room shapes, obstacles, and cold walls, aims to minimize the temperature gradient inside the room and places the pipes accordingly.

The available settings include the spacing of the pipes and distance from the walls. The entry and exit points for circuits can be freely defined by the user.

Automatic routing allows users to easily test different layout suggestions by adjusting the settings. Once a suitable layout has been generated users can install it to the selected room. The design can be further modified using standard MagiCAD functions.