In the prevailing COVID-19 pandemic, the concerned agencies/industry of the nation are making extreme efforts in providing the required amount of medical oxygen in hospitals. Moreover, wastage or leakage of oxygen in hospitals is advancing the challenges in the present situation. In this critical situation I made this project for minimizing the wastage/leakage and misuse of oxygen at all levels and ensure the best utilization of available oxygen for patients. 

This setup will be able to detect Oxygen Leakage Hotspot while the patient is sleeping with the oxygen mask on as improper fixing of oxygen mask on face is a reason for oxygen wastage . Besides , it can also be used while we are transferring liquid medical oxygen (LMO) from a capsule (Mobile tank) to tanks installed in hospitals. 

The idea struck me when one day I noticed that my household gas cylinder was leaking . During this it was making a ‘Hissing’ sound which was quite low but still audible to me and I inferred that if the leakage is too low it would at least generate an ultrasonic noise source that is distinguishable from mechanical noise sources. This brought me the idea of designing an acoustic camera that can be used to detect these noise sources by increasing the number of microphones and creating an array of them and then multiple layers of such an array of microphones to capture the leakage. The aim was thus to make the leakage (sound of leakage) visible through the coordination of camera and microphone . To capture this sound I thought of using a used Dish Antenna to place the large microphones in a manner to create a web of microphones increasing the sensitivity due to its concave shape as well.

The setup here consists of a camera, for capturing an image, a group of digital microphones ,and a Raspberry Pi Model 3 to read data from the digital microphones simultaneously and then transfer the data to a PC for post processing. Delay and sum beamforming is implemented in this design. 

3D Model of the Device in Isometric view

In this project, a PC is used as a main processing unit and a webcam is connected to the PC via USB port. A Raspberry Pi 3 is used as a data interface by reading data from the microphones and transferring the captured data to the pc .The signal acquisition and processing begins by making the webcam to take a snap shot photo of the hospital ward where the patients are sleeping or where oxygen is being transferred and then the image data sent directly to the PC via USB. A Raspberry Pi 3 is employed as a data interface device aiming to read

sampled sound from the microphones then stored the data in the onboard memory.

When the sampling process is finished, the data is then transferred to the PC for processing. A MAT LAB based

software with graphic user interface was developed to display and control the acoustic camera. The beamforming output are calculated then overlaid onto the image and, lastly, displayed as the acoustic image.

Hardware Connection Diagram

At the smaller scale sometimes due to improper fixing of the oxygen mask on the face, the oxygen is not properly utilized to support the patient’s needs. If the mask is not properly fixed on the face, some amount of oxygen is not inhaled by the patient and it is gone as waste. To compensate for the need of the patient, the flow rate of oxygen is increased which further increases the wastage of oxygen.

This solution, I believe, can solve for all levels in the current supply chain system of oxygen being transferred. 

Low Cost Frame for Making the Device Robust in a Hospital Ward

In this project, an acoustic array is designed and developed using a Raspberry Pi as the data interface module and digital microphones as the transducers. Thus, Raspberry Pi can be implemented to read data from the channels of digital microphone via GPIO port.


Leave a reply

Your email address will not be published. Required fields are marked *



Log in with your credentials


Forgot your details?


Create Account