Automated layout generation of HVAC piping from building floor plans

The design of HVAC water piping systems is often complex, time-consuming, and reliant on manual efforts. This paper investigates whether an automated method can generate pipeline layouts that meet functional, hydraulic, and spatial requirements. The proposed method applies an escape graph-based spatial modeling and a multi-agent ant colony algorithm to optimize pipe length and the number of local resistance components while avoiding obstacles, followed by pipe and valve sizing and hydraulic balancing to ensure system performance. Results show that the method can automatically produce feasible and efficient layouts while significantly reducing design time and manual workload. These findings highlight the potential of automation to improve both efficiency and quality of HVAC piping design in complex environments. Future work will focus on refining the fitness function, embedding hydraulic constraints into the optimization process, and enabling direct comparison with manually designed systems.