The laws concerning safety on worksite have become more strict year on year being that topic always more important. The main causes of injuries are workers unawareness and structural failures of machineries, i.e. cranes, or scaffolds. In order to prevent injuries, an accurate maintenance is mandatory. During the last years a new trend in maintenance has arisen in opposition to programmed maintenance. Monitoring the system, the condition based maintenance may prevent accidents avoiding unnecessary stops of the machineries and the corresponding reduced returns.
In this paper a health monitoring algorithm for a typology of construction machinery (the concrete displacing booms) is proposed. The proposed algorithm is based on the knowledge of geometrical and dynamical parameters of the boom, estimated through a stand-alone self-learning procedure. This feature makes the developed diagnostics system predisposed to be easily extended to other machineries or work fields. The common failure conditions, such as an overload or a crack propagation, are readily signaled.
The algorithm has been numerically and experimentally validated on a specific test rig which reproduces a reduced scale concrete displacing boom. The results, referred to the detection of two simulated common failure conditions, are presented and discussed.