ITWire

Winn's presentation had several core messages: to convey the enormous scale and complexity of the project and the potential for things to go pear shape if planning and execution are not meticulous; to communicate the depth and breadth of Telstra's expertise and the commitment made by it and its key suppliers to the project.

"This project on any measure is extremely complex and very risky." Winn said. "The executional capability to deliver this in the timeframe required and to do the work without disrupting commerce and public communications is enormous."

"You are playing with every working circuit in Australia - or 90 percent of them and you are going to have to test every one of those lines. We are going to have to visit every one of those premises and test back to the cabinet."

Winn suggested that there were many factors that could disrupt performance of the network: such as blenders and arc welders injecting interference into the copper whose ability to carry data is being pushed to the limits of technology. He envisaged the possibility of some future home electronic gadget, as yet unconceived, that could become enormously popular and create unanticipated disruption.

Winn claimed that Telstra had spent over three years and had dedicated the resources of 900 engineers to the project; a commitment that he claimed had been matched, or exceeded by its key supplier, Alcatel-Lucent, with no guarantee of gaining any business.

"Alcatel has 1.5 million person hours and a thousand engineers globally that have been dedicated to Telstra, to this architecture and the design and three years invested into this...There is no way that anyone who started in April is anywhere near where we are. The design is done, it is complete and we're ready to go. We have been ready for some time."

For Alcatel the gains are potentially enormous. Winn said that the network would require 50,000 kerb side cabinets: double the quantity Alcatel-Lucent has shipped to date, globally.

Winn also warned of the dangers inherent in sub-loop unbundling which, he said would destroy the integrity of the network and distribute responsibility for management and fault finding among multiple organisations

"Sub loop unbundling is the dumbest, stupidest thing I have every heard of - you will destroy the integrity of the network. The quality will degrade. And whose problem will it be?"