REMOTE MACHINE CONDITION MONITORING BASED ON WIRELESS CONNECTIVITY

Implementation of systems for diagnostics and monitoring of dynamical behavior allows identification of possible potential problems prior an equipment failure. Early machine monitoring systems determined the need of maintenance, however, modern systems also assist with severity assessment, avoid false positives, and help with process optimization. The subject of the present paper is an intelligent system for measurement of acceleration. It is based on micromechanical sensors with digital and/or analog outputs. The system is 8channel, allowing measurements in extended dynamic specter. The implementation is based on the low-cost microcontroller PIC18F422. A combination of measurement techniques, such as continuous self-calibration and multi-port measurement are used to obtain high accuracy and long-term stability. A data logging feature is also included allowing to store and lately analyze measurement results. Additionally, it is provided a connection with a PC for computing and visualizing the data. The software, realizing the communication and analysis is built as an upgrading and extending part of LabVIEW environment. The latter offers excellent possibilities to perform a narrow band spectrum and fractional-octave analysis of the vibration signals. 1. PREDICTIVE MAINTENANCE AND SENSORS Predictive maintenance method has several advantages over the other maintenance methods. Machinery can be assessed for maintenance while it is still running. The assessment uses non-invasive methods which are accurate and relatively inexpensive to install. From the running assessment maintenance, an action can be planed to take place at a time convenient to both machine and maintenance staff. The predictive maintenance requires continuous monitoring of the machines. This is achieved through the utilization of different sensing devices. The sensors are integrated within the machine system and give the required information on the actual condition of the system. Different types of sensors may be required to do this. For example, force sensors, pressure sensors, accelerometers and so on. Vibration is often the signal of interest for machine condition monitoring. Vibrations are usually caused because of unbalance and misalignment in moving parts of the mechanical devices. Left without correction, they lead to decrease of durability,