Condition monitoring of machinery in industries is becoming an emerging technique for efficient operations and productivity. A wide range of fault diagnosis approaches have been proposed to improve machinery operations in industries. These monitoring and fault diagnosis approaches are most effective for reliable machinery operations. However, due to the harsh environment of industries, these approaches have been suffered from different challenges. In order to extract the significant features for fault diagnosis and monitoring, different neural, fuzzy, and signal processing based systems were adopted. In this paper, we discuss wireless sensor based fault diagnosis and monitoring approaches and their types for industrial machinery. Furthermore, the paper presents industrial challenges to adopting these approaches and related efforts.