Neural Networks

Inspired by the human brain, artificial neural networks are designed to recognize patterns

Artificial neural networks, a component of machine learning, are a set of algorithms somewhat inspired by the human brain in that they are designed to recognize patterns. Basically, these computers or machines use examples to “learn” how to perform tasks instead of being programmed to do them.

A study in the International Journal of Architecture, Engineering, and Construction shows that artificial neural networks have been used successfully in many applications, including cost prediction, scheduling, risk assessment, claims, and dispute resolution outcomes and decision making.

For example, Natural Language Processing (NLP), which uses artificial neural networks and data science, is a useful tool in gauging public sentiment by processing hundreds of thousands of free text samples. The free text being information collected on social media, web-forms, and news feeds as opposed to formal survey data.

Imagine a computer that can eavesdrop on local conversations taking place all over the web and in a matter of seconds, recognize patterns in those conversations, and determine positive or negative sentiment. You might discover mass-transit riders are frustrated enough by trains running late to abandon the service altogether. City and town public works departments would be alerted to overflowing trash bins in a specific location and immediately take steps to solve the problem. 

In Australia, Ferrovial Services Asset Management has used NLP and artificial neural networks to help with rail asset management. In one example, 750,000 free text comments from maintenance technicians were analyzed to identify key issues and assets to focus on. There is simply no way humans could have found, reviewed, and gathered insight from that many records fast enough to actually benefit from the information.

Artificial neural networks are something our Digital Hub is studying to improve operations across all our services as we continually strive to find ways to improve operations on projects around the globe.