For the last two decades, developments in damage detection algorithms have greatly increased the potential for autonomous decisions about structural health. However, we are still struggling to build autonomous tools that can match the ability of a human to detect and localize the quantity of damage in structures. Therefore, there is a growing interest in merging the computational and cognitive concepts to improve the solution of structural health monitoring (SHM). The main object of this research is to apply the human-machine cooperative approach on a tower structure to detect damage. The cooperation approach includes haptic tools to create an appropriate collaboration between SHM sensor networks, statistical compression techniques and humans. Damage simulation in the structure is conducted by releasing some of the bolt loads. Accelerometers are bonded to various locations of the tower members to acquire the dynamic response of the structure. The obtained accelerometer results are encoded in three different ways to represent them as a haptic stimulus for the human subjects. Then, the participants are subjected to each of these stimuli to detect the bolt loosened damage in the tower. Results obtained from the human-machine cooperation demonstrate that the human subjects were able to recognize the damage with an accuracy of 88 +/- 20.21% and response time of 5.87 +/- 2.33 s. As a result, it is concluded that the currently developed human-machine cooperation SHM may provide a useful framework to interact with abstract entities such as data from a sensor network.