Non-Invasive Wireless IoT Oil Level Monitor Node using Submersible Sensors for Early Leak Detection

The early detection of safety critical conditions of oil levels in legacy industrial or commercial hydraulic machinery is becoming a new requirement for cloud-based maintenance service and Smart Grid systems. Building communities integrate accurate smart sensors to enable near or real-time monitoring. The nonideal conditions of oil levels can lead to disastrous costly events. However, legacy systems have gone expensive rigorous safety qualifications, to alter them makes the integration of new IoT monitors a difficult decision. To avoid or mitigate safety related recertification, two test setups are proposed to determine the sensor accuracy: (1) Full-range liquid height level test at room temperature and (2) Limited-range height level test at varying temperatures to characterize its nonlinearity effects on the sensor accuracy. During the empirical testing to obtain an optimal solution, the study proposes the use of Design for Six Sigma (DFSS) statistical method, an industry best-practice methodology to minimize parametric variations -by design- and increase the accuracy through measurements of variability and central tendency. Initially, the measured data drifted more than 50 percent from the desired values for both temperature and height readings. After the optimization, the temperature performance improved to a tighter specification of 0.25 percent from the initial prerequisite of 5 percent while the height performance improved from 8 percent starting requirement to 5 percent optimal specification. The DFSS empirical test is an effective method to achieve repeatable highly accurate oil level monitor.