One of the important challenges for biomass combustion in industrial applications is the fouling tendency and how it affects to the boiler performance. The classical approach for this question is to activate sootblowing cycles with different strategies to clean the boiler (one per shift, one each six hours,...). Nevertheless, it has been often reported no effect on boiler fouling or an excessive steam consumption for sootblowing. This paper illustrates the methodology and the application to select the adequate time for activating sootblowing in an industrial biomass boiler. The outcome is a control strategy developed with artificial intelligence (Neural Network and Fuzzy Logic Expert System) for optimizing the biomass boiler cleaning and maximizing heat transfer along the time. Results from an optimize sootblowing schedule show savings up to 12 GWh/year in the case-study biomass boiler. Extra steam generation produces an average increase of turbine power output of 3.5%. (C) 2009 Elsevier Ltd. All rights reserved.