Accurate measurements of pneumatically driven particle mass flow rate and particle size are necessary in order to maintain optimal combustion efficiency in coal-fired power plants and cement manufacturing facilities, as well as numerous other operations which are fed by multiple particle injection ports. Existing sensors for pneumatic particle concentration, such as concentration sensors, are typically prohibitively expensive and their measurements are sensitive to particle moisture content. A new sensor was developed for particle concentration measurement in pneumatic pipeline flow based on measurement of particle effects on fluid-induced oscillations of a probe extending into the flow. Since fluid-induced oscillations occur at a much lower frequency than do particle collisions, the measurements can be made with much simpler and less expensive equipment than is the case with impact sensors that attempt to resolve individual particle collisions. Experimental tests indicate three statistical measures of the probe acceleration data that exhibit smooth variation with the particle diameter and concentration data. The experimental data are used to train a neural net, which serves to interpolate the data. It is found that by training the neural net data on the statistical measures of the sensor probe acceleration, information on the particle concentration field can be extracted with a reasonable degree of accuracy.