{"id":3575,"date":"2025-11-25T11:13:19","date_gmt":"2025-11-25T10:13:19","guid":{"rendered":"https:\/\/neuraldesigner.com\/solution\/gas-emission-reduction\/"},"modified":"2025-11-27T12:06:44","modified_gmt":"2025-11-27T11:06:44","slug":"gas-emission-reduction","status":"publish","type":"solution","link":"https:\/\/www.neuraldesigner.com\/solutions\/gas-emission-reduction\/","title":{"rendered":"Gas emissions reduction using machine learning"},"content":{"rendered":"

Reducing gas emissions is one of the key issues industrial companies must address nowadays, both due to ethical necessity and new legal frameworks regulating pollution levels.<\/p>\n

It is often difficult to determine how to reduce these emissions, usually relying on trial and error.<\/p>\n

A machine learning model can help identify the actions to take to reduce the atmospheric pollution generated by an installation.<\/p>\n

This approach promises cost savings because it enables factories to anticipate the impact of changes in their control variables on gas emissions beforehand.<\/p>\n

Contents<\/h2>\n
    \n
  1. Objectives<\/a>.<\/li>\n
  2. Benefits<\/a>.<\/li>\n
  3. Approach<\/a>.<\/li>\n
  4. Conclusions<\/a>.<\/li>\n<\/ol>\n<\/section>\n

    <\/p>\n

    \n

    Objectives<\/h2>\n

    Gas emissions reduction attempts to model atmospheric pollution, allowing for an understanding of its dependency on control variables.<\/p>\n

    <\/p>\n

    The challenge is to get information on variables that can be easily controlled without decreasing the factories’ profit.<\/p>\n<\/section>\n

    <\/p>\n

    \n

    Benefits<\/h2>\n

    It enables making optimal adjustments to the factories’ control variables to minimize gas emissions.<\/p>\n\n\n\n
    <\/p>\n

    REDUCE DOWNTIME<\/h3>\n

    <\/p>\n

    LOWER COSTS<\/h3>\n

    <\/p>\n

    REDUCE GAS EMISSIONS<\/h3>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/section>\n

    <\/p>\n

    \n

    Approach<\/h2>\n

    The way to reduce gas emissions is to create an approximation model that considers input variables related to our target and estimates the latter.<\/p>\n

    Neural networks can model the correct values of the target variable to know its dependency.<\/p>\n

    That saves costs and time on decision-making and reduces a factory’s emissions.<\/p>\n<\/section>\n

    \n

    Data set<\/h2>\n

    The data set<\/a> contains measurements from our system or process.<\/p>\n

    It comprises state, control, and performance\u00a0variables<\/a>.<\/p>\n

    <\/p>\n

    State variables<\/b> are those\u00a0inputs<\/a>\u00a0that determine the system’s performance and are not actionable by the company’s technicians.<\/p>\n

    Some examples of state variables are:<\/p>\n