This U.S. Bureau of Mines report presents results from an ongoing investigation of the use of adaptive signal discriminating methods to distinguish between cutting coal and cutting mine rock. Bit- cutting forces and tool vibration were measured in the laboratory as a linear-cutting apparatus (lca) made constant-depth cuts in coal, sandstone, and shale test specimens. A portion of the collected data has been analyzed, and some preliminary results are given here. The influence of data bandwidth, data window size, number of signal features, and voting among classifiers on classification performance are noted. Results to date, based on ideal cutting conditions and simple geologic materials, indicate that of the four classifiers tested there appears to be no single best classifier. Usually, classification accuracy showed slight improvement as the number of features considered for classification increased. The highest classification accuracies were achieved when voting was conducted among classifiers followed by voting among force components.