Low-power wireless accelerometer-based system for wear detection of bandsaw blades

The paper provides a framework to save energy and reduce the operative cost of some of today's industrial machinery. Low cost and low power wireless sensor networks is a novel approach to monitoring the tools in order to save energy and keep the tools monitored. Cutting tool wear degrades the product quality in manufacturing processes and also could have implications in health and safety of use. Monitoring tool wear value online is therefore needed to prevent degradation in machine quality. Unfortunately there is no direct way of measuring the tool wear online which is also very low cost. In this work is presented a low power and low cost accelerometer-based system for wear detection of bandsaw blade. The algorithm uses a simple data processing directly on board that can extract features and perform a classification on the state of the blade. Low power design of the node, on board processing and wake up radio capabilities reduce the wireless communication and the power consumption of the node significantly. Experimental results show the high accuracy, up to 100%, of the algorithm and the low power of the proposed approach.