Controlling EVM indices in construction projects by a statistical

Earned Value Management (EVM) is considered as a well-known approach in project control system. Although the EVM indicators are the most effective tools in the measurement of project cost and schedule performance, they cannot show continual performance variations. Considering the EVM indices, Statistical Process Control (SPC) methods can monitor variation of project performances persistently, and they lead to improve substantially project controlling scheme. In this paper, a selfstarting monitoring using multivariate control chart is applied in the analysis of earn value indices owing to costly sampling and lack of historical data in the projects. Also, these indices are assumed auto-correlated to enhance more applicability in real cases. In addition, usually EVM indices are distributed non-Normal in practice. For this end, auto-correlated earn value indices are monitored with considering correlation between them in which indices can be distributed non-Normal. So, a Mont Carlo simulation study is done regarding the probability of the signal criterion. Finally, the performance of the proposed approach is shown through an illustrative example.

Correlated schedule and cost performance indices; Self-starting monitoring; Autocorrelated non-Normal EVM indices