Background Research

Vibration sensors have extensive applications in predictive maintenance appliactions and have been used in the industry for long enough now. However, not many applications have found ground outside the manufacturing sector. A google search for projects on vibration ssensors returns pretty timid results like shock detection or alarm detection and so on. However, the industrial grade sensors from Keemet offered in this roadtest do not deserve to be used in such silly applications. The wide bandwidth and high noise rsistance the sensor provides makes it ideal for a serious predictive maintenance project in the industry - but wait - I just said that we are not doing that

And yes - we aren`t - because I have much more creative applications in mind which I would like to use these for.




I was startled by the precision rating of the vibration sensor when I read the datasheet and on the basis of that thought of some applications which I  would like to implement. The ideas are as follows -

- To measure natural vibration in an IC engine vehicle to calculate the precise accelaration and other IMU readings. The need for this project arose from an industry project I was working On in which we had to control a driverless IC engine based car based on IMU readings. The problem was that due to the vibration of the engine, the IMU readings turned out to be erroneous and so was the vehicle control. Obviously this difficulty was offset with an optical flow sensor but it required additional computaion due to sensor fusion and all. We have tried many vibration sensors with low precision to not much avail. However, this sensor semms to be the only piece of hardware we were missing.

I really would like to work on this project ASAP but the problem is that the Desgign challenge deadline is in the next two months while there seems to be only a faint chance that my university would allow us to return to  campus before that. Hence, I would not like to take a chance and keep this project on hold for the time being.

- The second idea I have which is  probably a more generic one that can be applied to many situations is the predcitive scheduling of maintenance tasks for household appliances. Being more specific, I would like to create a system that monitors the vibrations in the compressor of an air conditioner to predict when it is time to clean the air filter mesh. Research proves and I have practically experienced that the vibrations in the compresor bears a direct correlation to the filtering coefficent of the mesh, ie in layman terms, how dirty the mesh is or how much dust particles are clinging onto the mesh. MAny automatic cleaning air conditioners rely on open loop cleaning of the mesh at periodic intervals. this however is a power hungry process as micro motors have to be actuated to move the cleaning head over the mesh. Also, in air conditioners that do not have the automatic cleaning feature, manual intervention is required at regular intervals. Since, the life of an AC is affected by its compressor`s load cycles it would therfore be a welcome investment to install predictive maintenance feature in it. This could translate to huge savings in time and money spent by corporations in maintenance of air conditioners in large buildings and laboratories, ie when the count of such machines is large

The concept may be extended  to other types of appliances like refrigerators and heaters or even air  purifiers. The reason I do not mention them here are that firstly, I did not get any research paper backing the correlation of vibration intensity with the life of these particular appliances and the fact that it is much easier to simply install this in my air conditioner at home




The final solution to the above problem is a vibration sensor capturing the vibrations in the compressor of an air conditioner feeding raw data to a MCU which processes thue data and sends warnings and maintenance alerts to relevant stakeholders. The notifications wuld be issued in two ways. Over BLE, the warning is issued if the user or owner of the appliance is detected to be nearby(by seeking the bluetooth connection on the user`s devices like laptop or phone). The MCU would also be interfaced with anothuer MCU capable of Wifi connection so that maintenance information can be stored on the cloud and alerts can be received remotely.




The sensor would be mounted directly onto the compressor with heat resistant tape(the one used to stick heat sinks - I have one from my raspi collection). The RAW data is feed to one of many analog channels on the STM32 Nucleo development board. After filtering the signal using FFT based DSP algorithms available as a ready to use API on the STM32 CUbe Studio, the signal is fed into a spiking neural network that analyses, characterises and quantifies the vibration. The output of the spiking neural network is a command to the MCU to either issue a maintenance request or not. The MCU acting on this notification issues a warning to the user over bluetooth with the help of the shield provided if the user`s devices are in range. If not, the MCU sends a message to the ESP12F module conneted to it over UART. Before this however, it sends a wake-up notification to thue ESP12F module to wake it up from sleep mode. Once 'awake', the ESP module connects with the home Wifi network and sends a notification over MQTT to the Ubidots Cloud platform. A webservice running online then issues a message to the user to take necessary action. The STM32 board and the ESP MCU would be housed in a plastic case attached to the AC and would continuously monitor the state of the appliance. The apparatus would be powered from a regulated 5V/1A AC to DC power supply.



My Background

I am a third year undergraduate student at the Indian Institute of Science, Bengaluru, India, enrolled in the Department of Microelectronics. Throughout my academic tenure at this leading research institute, i have worked on many electornics and embedded projects in industry and research oriented applications. I usually spend my time working on some electronics project in our labs since our institute promotes 24x7 access to research facilities and when I am not doing that I am  either busy with personal hobby projects.

I am new to element14 and have no idea why it took me this long to stay unaware of something this cool. Seems like this community is exactly what I was looking for to diplay all my hobby projects and learn from veterans. This competition I believe is the perfect starting point of my journey in this brilliant community.

Talking about my skillset, I am skilled in signal processing, machine learning, robotics and embedded system applications in these fields. My recent projects include a Brain Machine Inertface to control a Robotic hand and an autonomous indoor bot. I learnt to apply the nitty grities of signal processing in this project. Since, the project I propose is somewhat similar, it would be great to do something that would be used in real life - outside the laboratory. I have also worked on several robotics projects in the past. I recently developed an autonomous robot for delivery of essentials durig the pandemic.

As part of my internship at an automotive firm, I worked on the embedded systems design for a driverless vehicle I talked about earlier. It was a really cool project in which I learnt a lot about sensor fusion and noise elimination from sensor readings. Apart from these I have done several hobby projects centered around IoT, robotics and basic electronics(discrete circuits).

Hence, I believe I am well-equipped with the necessary skills to complete and probably win this challenge. I can commit to give an hour everyday towards this project. I know I am new here but I think this is the best chance I have to get involved and contribute to the community.