Refraction, the method by which light waves can be guided, was first demonstrated in the 1840s by Jean-Danialel Colladon. Twelve years later, John Tyndall included a demonstration of the effect in his public lectures in London. It wasn't until over 100 years later, in the 1960s, that optical fibres were proved to be useful as a method of communication.
Since then optical fibre has been primarily used in long-distance communications, due to its considerable advantages over copper wires. As prices have fallen however, more and more fibre is being rolled out to the home. 99% of all data worldwide is now transmitted over optical fibre.
With the increasing world population and increasing numbers of connected devices with the growth of the "internet of things", data traffic is growing at an ever-increasing rate. The growth in capacity of optical fibre cannot keep up with demand and it is predicted that by 2020 demand will outstrip capacity. This is commonly referred to as the "capacity crunch". If this happens, bandwidth will be rationed and prices will increase. This is why the TRANSNET team are foraying into Machine Learning as a basis for their research, to ensure the infrastructure of a rapidly expanding ‘internet of things’ is sufficient.