Applications in Secondary Sector

Machine Learning applications in Secondary Sector-Part 2

Posted on Posted in Machine Learning

This blog includes machine learning applications in Secondary Sector which is a continuation of the last blog post.

Detailed below are some more machine learning applications in Secondary Sector:

Construction

Construction, engineering and management use artificial neural networks through predicting risk and optimization, classification and choice of resources. Some examples are training neural networks to study basic soil properties from a dataset of different cohesive soils from various location and solving engineering design and image recognition by pattern recognition .

Food Processing Industry

An example of machine learning in food processing industry is broiler industry in Iran. Iran is one of the top 10 producers of chicken meat in the world and uses machine learning for making models of important environmental trends such as prices of maize, soybean to predict the cost of chicken meat and product supply chain. The trade ministry regulates the market in case of deviations to give food and save jobs.

Shipbuilding

Machine learning techniques such as non-linear optimization, hierarchical cluster analysis and principal part analysis estimate costs, for example, The North Atlantic Treaty Organisation (NATO) estimated the cost of amphibious assault ships in Netherland .

Breweries

Due to the process involving a use of living organisms and biological raw materials it is difficult to set up control in drink’s industry. Daily measurements assist monitoring the process and keep it to a desired course of fermentation.  Neural networks provide a mechanism to help monitor and control fermentation. Also, reliable predictors lead to fewer measurements and early warnings to make corrective actions. The relationship between flavours and sensory perception model predict flavour of certain types of beer.

Energy Utilities

Machine learning can help demand side management by reducing computation time and predicting energy consumption. Networks identify faults through pattern recognition and efficient data analysis. It also helps in diagnosing performance by identifying issues in electrical equipment and address fluctuation in power generation. An example using machine learning is a California-based firm called App Orchard which performs analysis using simple natural language query to manage transmission and workforce.

Metal Working

Artificial neural networks help manufacturability exploration, process planning, and die design. Finite element methods find out the non-linear relationship between design and input parameters. Analysis and modelling of metal-cutting process determine ideal response values and, therefore, increasing the product’s quality and life of the tools being used.

 

Leave a Reply

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