This study quantitatively measures the solid flow properties using an optical fiber probe (OFP) in a 0.105 m diameter cylindrical lab-scale gas–solid fluidized bed. The influence of superficial gas velocities and static bed heights on particles of different size ranges of Geldart B classification (100–200 µm, 200–300 µm, and 400–600 µm) are systematically analyzed. An increment in the local and cross-sectional mean solids holdup is observed with the column static bed height along with the axial and radial directions. A reduction in the cross-sectional mean solids holdup is noted with the superficial gas velocity at different axial positions. It is observed that the local solids holdup is greater near the wall than at the center of the column along the radial direction. With the increase in particle size, a higher value of both local and cross-sectional mean solids holdup is noticed at different axial positions. In addition, various statistical tools such as the intermittency index, radial non-uniformity index, and kurtosis are used to analyze the solids distribution behavior inside the fluidized bed. The results showed an increased value of non-uniformity in voidage with higher gas velocities. Furthermore, the intermittency value indicates higher fluctuations in the bed center than the wall. Finally, the artificial neural network (ANN) model is developed for solids holdup with five input variables. A single hidden layer with nine neurons accurately predicted the solids holdup. © 2022, The Author(s) under exclusive licence to Associação Brasileira de Engenharia Química.