Had a great chat this morning with Cerion’s Operations Research (OR) guru – Dr. Wei Yuan.

Wei helps Cerion to develop mathematical simulations for the most complex and challenging real world problems and then uses algorithms to make better decisions, helping our customers find the best possible solution to any given problem. 

Much of Wei’s time at Cerion has been spent developing the optimization engine that we have built into our mobile network engineering product to improve the performance of these networks and make them more efficient. Indeed, since joining the company, Wei has filed several patents in this area as well as that of making the ongoing configuration management of networks more efficient.

Considering the general increasing complexity of networks and the ever more competitive wireless marketplace – Wei and I share the view that Operations Research (OR) is something the industry needs to leverage more and more, to remain competitive and be successful. We have a tremendous number of ideas about where OR could be applied &/or used on a broader scale, to help reduce costs, boost revenues and generally make our customers internal processes far more efficient. 

Recently this has led us dig far deeper into artificial intelligence – An area where we are already using learning algorithms to help characterize the behavior of wireless networks and customers, to make better engineering, marketing and other decisions.  Interestingly many of the methods we are using in this area are based on ideas that go back as far as the seventeenth and eighteenth century, devised by mathematicians such as Thomas Bayes and Pierre-Simon Laplace. 

With the advent of ever-more powerful computing capabilities and advances in highly scalable fast-access databases, it is only relatively recently that it has been possible to apply these ideas to help with machine learning and the more general area of artificial intelligence. 

Focused on the transaction analysis that I have discussed in previously Cerion is applying artificial intelligence to help simulate, learn and charachterize the true capacity of the different  mobile network elements. By linking this with a forward trending capability and various optimal configuration algorithms we have already developed, we are enabling our customers to pro-actively understand exactly how much equipment they will need in their network, as well as where and when.

One of our visions when we founded Cerion was to devise solutions that would help networks become self provisioning - And now leveraging a combination of our own ideas and those of the master mathematicians of the past, it seems we are well on our way to making this a reality.

Peter