IoT Integrated Shaft Condition Monitoring System

What it does

Health condition monitoring system integrating Internet of Things, mechanical technology and signal processing to predict the failure of automotive power shafts improving safety and driving conditions.

Your inspiration

I always believed that preventive measures are better that solutions to problems. Hearing of an accident involving 12 vehicles on Highway 427 in Toronto due to the failure of a vehicle's power shaft lead me to condition monitoring system. With an aim to prevent such accidents from occurring, I took a keen interest into the working of preventive engineering. Similarly to how earthquakes are predicted thanks to the vibration in the ground, I applied the similar concept to the predictive model with an addition of an IoT focused system.

How it works

Vibrations are picked up by an accelerometer (acceleration sensor) of high accuracy. These vibrations are sent through a data acquisition unit consisting of a DEWE-43 and MyRio devices to convert the analog signals to digital for processing. Signals of the shaft are sent to the Raspberry Pi, which is installed along with devices for computation that is prepared to perform localized decisions when a problem detected. Once the detection reaches the threshold limit, the information will be transferred to the Internet of things (IoT) connected with a ground control centre. If no response received from the control centre within the predetermined period, the fog decision system will execute the predefined decisions to save the automotive as shown in the images (Health Conditioning Monitoring Framework) provided. A flowchart depicting the relationship between each system is located in the images provided.

Design process

Rotary bending analysis (physical simulation mode to study the effective power transfer in a rotating shaft) of the specimen Aluminum alloy AA7475 under similar supported conditions of an automotive power shaft. Experimental set-up at a laboratory scale to run rotary bending fatigue test can be seen in the images provided. The experimental set-up and test configurations were based on the recommendations by ASTM E2948-22. Specimens consisting of 6 and 12 mm neck diameters were subjected to loads of 20, 30, 35 and 40 N at 650 RPM till fatigue fracture. Specimens were manufactured in a CNC T20 to ensure uniformity. Electron imagining was carried to identify fatigue patterns on the plane of failure of each specimen as shown in the images provided. Vibration data collected by the data acquisition unit was processed by Wavelet and a Fast Fourier Transform algorithm (segregated signals into pure sinusoidal signals). The processed signals are used in the health condition monitoring system to condition of the power shaft. Bluemix is used as the IoT Cloud Integrated framework for its simplicity and efficiency. To validate the results obtained from the experiment both the rotary bending analysis results and experimental results were compared.

How it is different

IoT Integrated Power Shaft Condition Monitoring System can be considered as a unique system as there are non-like it on the market. Combining both IoT and condition monitoring techniques and practices, a prediction model of a transmission failure was developed. Condition monitoring smart devices are indeed available in the market but there are no developed predictive models for automotive transmission failures or an IoT connected system between condition monitoring devices and control centres.

Future plans

I intend to expand the laboratory scale setup to include a transmission power shaft as the specimen replacing the Aluminum alloy. Developing a supervised machine learning program to improve the prediction model of transmission power shaft. I intend to further develop the system by incorporating more compact and connected devices to reduce its weight and size restrictions. I want to have this system become a standard in the design of vehicles similar to that of the condition monitoring system of the engine (engine light).

Awards