Wireless Sensing for Ground Engaging Tools

Thumbnail Image
Ibrahim, Adel
sensor node , sink , hardware design , ground engaging tool , shovel teeth , proximity sensor , harsh environment , entropy , entropic sensing , entropic filter , wireless sensor network , internet of things , industrial internet of things , time series , energy efficiency , Shannon's entropy
A long-standing industrial problem is tackled with an industrial internet of things (IIoT) application. This thesis presents the hardware design and development of a wireless sensor network for monitoring ground engaging tools of earth processing equipment. This work addresses several challenges in the hardware design. The challenges included establishing wireless communications from within a metallic enclosure, Investigating the effect of antenna tuning on the link budget, designing a protective package for the sensor node (SN) to withstand vibrations and mechanical shocks, and creating a low-power design that allowed for a minimum battery life of one year. Furthermore, this work puts forward the design of novel application-specific sensors including a low-power ferrous-selective proximity sensor and a low-cost capacitive wear-level sensor. Hardware designs were evaluated with laboratory and field tests. The conducted tests demonstrated the SN’s functionality and performance in harsh operating conditions. The journey from problem definition and to successful field testing of the complete wireless sensing solution is covered in this thesis. In addition, a novel energy-efficient approach to wireless real-time sensing is presented. For an SN transmitting samples of a discrete time-series in real-time, its lifetime depends primarily on its battery capacity. With most of the energy consumed in wireless transmission, this thesis introduces an energy-efficient scheme that significantly reduces the number of transmitted samples, while maintaining a low mean absolute error between the original and the recovered signal. The concept of instantaneous entropy is introduced and a computationally-efficient iterative formula for computing Shannon’s entropy is derived. The SN evaluates the information content in each sample and decide whether to transmit or omit the sample. At the sink, incremental machine learning is used to recover the omitted samples in real-time. This approach shows an average of 60% reduction in energy consumption by the SN with less than 2% mean absolute error in the recovered signal. The simulation employed stationary, non-stationary, chaotic, and random time-series for performance evaluation and benchmarking.
External DOI