AI-based Optimal ACH Flow Prediction to Reduce the Contaminants

Client Profile

The client is an established engineering and management consulting firm based in California with a background in developing healthcare solutions.

The client wanted a solution to predict the Optimal ACH (Air Changes per Hour) value of the airflow, from the pre-trained data to reduce the airborne particles in the surgery room that could infect the healthcare workers.
The client also wanted a 3D model rendering of the rooms, so that the position of the inlet and outlet can be determined along with the volume of the room.
The client wanted the whole operation to be performed on the cloud along with following HIPAA compliance.

Seaflux created the AI-based platform and trained the ML model using Artificial Neural Network (ANN) and simulation results of OpenFOAM as training data to predict the Optimal ACH value.
Seaflux developed the solution using AWS, where the users can enter the desired particle size, position of light, temperature, type of surgery, and outlet position in a 3D model to get the optimum value of the ACH.
The ANN algorithm would measure the inputs, compare them with the current surgery room status, provide the ACH value, and set the airflow of the room.
The AI model would keep measuring variables like temperature, particle size, and light to set the ACH value to the optimum capacity to remove any contaminants.

With the optimal ACH value calculated, there have been reportedly fewer cases, a 12% reduction, of the healthcare workers being sick.
As with the proper data of optimal ACH value, the power consumption of the machine has been reduced by 37%, in turn saving the operation costs.