Monday 8 April 2013

NBN: Turnbull pre-releases Policy to News Ltd, engages Full-FUD drive.

News Ltd is spruiking for the Coalition again, spreading Fear, Uncertainty and Doubt far and wide.

It's complete B/S for anyone, including NBN Co and Conroy, to say they know the exact project cost or completion schedule, either "on-time, on-budget" or a gross exaggeration of 2-3 times over budget, or even as Turnbull has been claiming, "it is a failing project".

I'm sorry, but "Failing projects" don't deliver anything, the NBN Co work is progressing and progressing well. The ALP and NBN Co are doing a lousy job of selling their performance and informing the public of their progress. MT would be right to criticise there.


Right now, the NBN project is in "phase two" (my term), rolling-out the necessary, but invisible, infrastructure. Just like building a house, you spend 30% of your money coming up to ground-level, apparently "doing nothing". It's much more if you count the cost of services that are buried & invisible: roads, drains, curbing, water, storm-water and sewerage - and electricity, phone and network now.

Early delays of a project may or may not affect the final schedule and cost. But what you can be absolutely certain of, if you don't rectify faults, errors and omissions in the foundations, you'll pay for it many times over later on or even cause real project 'failure'. Early foundational work is critical to project success, is essential to corporate learning and Adapting the Plan. As the US Marines say: Improvise, Adapt, Overcome... The Plan is not the built & delivered items, it's just paperwork, necessary and vital though, and shouldn't be confused with the work. Also, projects have a lot of overheads - things that are necessary and don't show up in the final deliverable. Much like the launch tower of a rocket, slipway of a ship or cranes and scaffolding of a building: you gotta have them to build the project, but they are "deliverables". The extreme case is tunnel boring machines: they cost $100's millions and are often buried in a dead-end once the main tunnel is finished.

The one certainty of Project Management:
The Plan you start with is not the one you complete the project by. Things change, you discover problems and invent better solutions.
The "project management experts" who are projecting forward current completion rates from Phases 1 and 2 to Phase 3 are either incompetent, ignorant or being mischievous and disingenuous. How quickly you build the Foundations doesn't predict the landscaping, interior decorating or "final fix". These calculations are as invalid and falsely based as the "risk assessments" on CDA's/CDS's that led to the US "sub-prime mortgage" collapse and GFC. Just because you plug numbers into a spreadsheet does not make them real! It's Garbage IN-Garbage OUT writ large for purely political FUD.

In 2014, the NBN project is scheduled to move into high-gear with simultaneous rollouts all around the country: because all $2 billion of common infrastructure that's been needed will have been designed, built and tested now.
"Phase  three", (my term), starts where the number of premises passed and connected will be measurable and relevant, and then the nay sayers can then rightly start criticising the roll-out.

Until then, Quigley and Conroy should start telling what real progress they're making so some of the FUD might be dispelled.

On other topics:

Turnbull has said he'll guarantee 25Mbps. That least that's something.
What he isn't saying is what proportion of premises will be deemed "cost-effective" under his plan, nor what service they will get.

Is the Big Reveal for tomorrow a rerun of the OPEL contract? WiMax or similar everywhere?

Also he'll keep the unprofitable $9.5 billion Fixed Wireless and Satellite parts of the NBN Co network: 25% of the cost to cover 7% of the population. It was always meant to be subsidised, but where's the income to do that going to come from if he shuts down the Fibre rollout or privatises the lucrative portion? Are we back to USO craziness or "Direct Action" payments from the Federal budget? Both are dodgy and unsustainable plans.

In the 2007 Telstra made to the ACCC on the G9/Terria FANOC proposal, they said Alcatel could provide them with VDSL2 systems (ISAM's?) that would do 25Mbps @ 1500m.

They also said they'd planned to rearchitect the copper network, optimising it for the new system (diagram is a bunch of cells), needing 40% fewer nodes than the FANOC 'leave the copper alone' design.

For 'the big 5' cities (& ~4M services) 20,000 nodes in the field [160 services/node] + 1.6M services directly connected to 450 exchanges.

It has to scale up to either 8.6M or 12.2M services... So that's 40-60,000 nodes plus 20-30,000 ISAM's in exchanges.

Alcatel is also quoted by Telstra [Telstra_2007_Triple+Play+Bandwidth+Requirements]
  • currently 25Mbps is needed for 'ordinary' household watching video etc, 8Mbps for HD video, not the 6Mbps Turnbull is now quoting.
  • but they project 70Mbps required in just 15 years... How's he going to pull off that trick?
  • upgrading the last-mile from copper to fibre will take the whole project cost significantly over doing full-fibre now.
To quote Telstra: Do it Right, Do it Once.

Why go to an interim solution when we all know that we want to go to full-fibre eventually?
Turnbull's side-step is: "In 20 years, there may be all sorts of new options invented that mean we don't go fibre".
Nope, absolute garbage: if it's not in the laboratories now, then it won't be in production in 20 years.

All the work on ultra-high-speed networks is in fibre-optics. Yes, there will be faster radio systems and faster copper network solutions, but they will fall further and further behind in speed.


The $100B question: will the independent regulator (ACCC) agree to the Turnbull/Fletcher plan?
Trujillo/Telstra spent 12-18 months to not reach an arrangement in 2005/6.

Turnbull can spruik his "great plan" and spread FUD deep and wide like manure all he wants.
Until he has signed contracts and regulator approval, it's just blowing smoke.

Turnbull hasn't done commercial & regulatory negotiations like this before, why does he think he can make it happen in a trice, or get a good deal? He's kidding himself and will be taken for a ride.


Other Unanswered questions:
  • What direct coverage will the Coalition give? 97%, 66%, 40%?
    • Leaving out the expensive parts of the network, "cherry-picking", is a simple cheat to get notional prices down, but will come back to bite him very badly.
  • Who will supply the Customer Premises Equipment or NTU? It's around 10-20% of per-premises cost.
    • It's part of the NBN Co costing, but never appeared in the Telstra docs.
    • Turnbull is now saying "no install will be needed in homes", so just how do subscribers connect to his FTTN and who pays for the device and its installation? The Network Fairy?
  • How is Turnbull doing telephony?
    • The Telstra plan was VoIP with ATA's in the node. The FANOC design was very poor.
  • What happens to the already built/contract NBN Co Fibre facilities (like PoI's)?
    • all the transit networks, backhaul, dark-fibre, ... 
      • How will they NOT waste the $2B investment?
  • Will all the various suppliers provide a single, integrated & transparent 'bitstream' service? (L2 ethernet).
    • i.e. will access seekers/RSP's still get to deal administratively and technically with a single infrastructure?
  • Trujillo is on record in 2008 that it would cost $15 billion to build his FTTN covering 98% of premises: only 50% of the $28 billion for NBN Co's Fibre.
    • Does Turnbull have better costings? From whom?
    • How can be claim his FTTN will be "only 25% of the cost of fibre"? It's rubbish.
  • What economic lifetime is Turnbull projecting for the new FTTN build?
    • Telstra had 20 years.
    • Will we be able
  • What Traffic Growth Model is Turnbull using?
    • ABS says 53% CAGR [compound annual growth rate]
    • CISCO VNI 30% CAGR.
    • Over a decade, ABS growth is 70 times (70,000%!)
    • CISCO is 15 times in a decade.
  • Which of those demand models does he favour?
    • Or has he invented his own?
  • Over two decades:
    • 30% CAGR is 225 times growth in traffic. [Or 1000 times if its 50% CAGR then 30%]
  • Just when does Turnbull think Fibre from Node to Premises will be needed?
    • Will we have to wait a full 20 years for it to be "Cost-Effective", no matter what?
    • My calcs show the fibre upgrade at a minimum of $12B + node upgrades + NTU's
    • Nodes + Fibre is the most expensive and slowest way to get to full-fibre.
    • If Turnbull doesn't organise a scheduled, full-coverage upgrade to fibre, it'll cost $3-5,000 for the first premise in an area to connect.
    • Creating a massive barrier to upgrade that all but a very few will afford, locking us into a worse information-divide.


No comments:

Post a Comment

Note: only a member of this blog may post a comment.