Steel structures are robust and enduring, however susceptible to various damages including loss of the grips of bolts at connections. Such secondary damage is difficult to monitor manually. With the help of applying smart computing techniques for structural health monitoring methods, namely use of Piezoelectric sensors (PZT) sensor by the Electromechanical Impedance Technique the monitoring of a bolt’s grip becomes effective and efficient. This paper represents results of experimental investigations on a bolted steel connection specimen prepared and tested for the damage detection conditions. The study focuses on identifying the response due to looseness with and without applied axial tensile load using different statistical damage indexes like Root Mean Square Deviation (RMSD) and Mean Absolute Percent Deviation (MAPD). The damage assessment is analyzed based on the signature of impedance concerning frequency, variation of torque in the bolts, and axial loading. Using the different number of repetitive experimental output data, the feasibility and effectiveness of this method are verified. The results show that both RMSD and MAPD damage indices increase with the decrease in the design torque in the bolt establishing an effective way of monitoring of the bolts’ grip. The looseness indices were found to be highly sensitive to small damage which showed early warning for repair or maintenance helpful for monitoring process. The results showed good agreement on application of smart computing for structural health monitoring using sensor-based damage detection necessary for the smart infrastructures.