Abstract: The amount of construction waste(CW) is huge and the composition is complex. If not properly treated, the asphalt, gypsum, heavy metals and paint in CW would react with the surrounding air, soil and water to produce harmful products, which would seriously endanger the human living environment. The construction requirement of “waste-free city” in China clearly put forward the strategy of comprehensive management of solid waste to improve the resource utilization rate of waste. As the largest municipal solid waste in China, CW has become the focus of attention, and it was urgent to realize the resource utilization of CW. The traditional CW disposal methods adopt manual and mechanical means to sort CW, and then make resource utilization of CW. There were some problems in the process, such as difficult management, low efficiency and high cost. As the core of artificial intelligence, machine learning has been gradually applied to CW treatment, which could effectively improve the resource utilization rate of CW. The basic situation of CW and machine learning were described, the process of CW treatment was introduced, and the research progress of machine learning in the field of CW treatment was summarized. Finally, combined with the national condition of China, some suggestions for CW treatment were given, in order to provide reference for realizing automation and intelligence of CW treatment.

Key words: machine learning, construction waste, artificial intelligence, resource utilization