Typically, when a production system goes down, only a small fraction of the downtime is spent repairing the machine that causes the failure. The greatest part of downtime is spent locating the source of the problem. Therefore, the development of in-process monitoring of machine degradation and fault is one of the most important research tasks for increasing machine uptime and improving production quality. This paper examines various methods in machine monitoring, fault detection and fault diagnostics and proposes a frame for autonomous maintenance management system. In the case study, a new approach is proposed for the on-line measurement of the maximum peak-to valley roughness of a finished-fumed surface in the feed direction. The method is based on determining threshold by using a linear least-square estimate of the angular scattered light pattern reflected from a surface and the in-process monitoring and control approach is also presented.