Thursday, 2 May 2013

NBN: Both Politically defined solutions are wrong.

It's trivial to show that both NBN designs are sub-optimal: both proposals could be much better, but they each side is locked into "Our Way or Nothing".

There is massive risk, upside and downside, in constructing an NBN stemming from:
  • unknown consumer demand and demand elasticity, and
  • unknown consumer willingness-to-pay.
To personalise the problem:
What if you threw a massive party, say a wedding, and nobody came, or everybody comes? Either way, you're in big trouble.
Neither of the Politically motivated proposals have a Plan B:
  • What if we get 40% through and nobody will pay for it?
  • What if we get swamped right from the start?
Neither side is able or willing to say just how they'll alter their plans when they hit Reality.
"Death by Success" is as problematic as "Failure to sell". 

But it's a little more subtle:
There are important edge cases that are completely ignored and both sides brook no change or comment to their Precious.
The people who have the most to gain from the NBN are, ironically, the least well-served: those in 'the bush', on rural properties or in small country towns. These folk also have the lowest average incomes in Australia.

They get to substitute remote access for "driving to town", long, arduous and dangerous trips, but also expensive. Ever complained on your holidays about the pump-prices? They get to pay those every day and on their lower average-earnings and 100,000+ km/year travel, transport is a much higher proportion of their budgets.

  • Both sides are pretending the NBN upgrade is a Free Lunch.
    • Neither side is willing to test what people will pay.
    • The Coalition has hidden quite a few nasty surprises in its little package.
      • When consumers experience the sting-in-the-tail, they won't be that impressed by a 10% saving on total cost.
  • Smaller towns may be much better suited to what Google are rolling out, active ethernet, at least initially.
    • Or a variation of the TransACT solution:
      • 100Mbps cat-5 ethernet for the last 100-300m.
      • It's cheap, it's DIY and it might last happily 10 years.
    • Small towns can provide "sweat-equity", like digging trenches and laying 100mm conduit.
  • Every rural property with an 11KV electricity supply could have a fibre service, at a reasonable cost, either GPON or active ethernet.
    •  ~$10,000/km for single 'tube' (12*12 fibres) on the power poles.
    • Properties on same spur-lines can share the cost.
    • Many of those properties are businesses, they are ready and willing to pay for fibre, if only it was offered.
  • One of the blockers for towns of 1,000 premises or fewer is the cost of the transit network uplink.
    • Currently for GPON, this is assumed to be fibre and $25-$35,000/km.
    • If towns are "off-route" they don't get connected.
    • Australia is known for its extreme Innovation, especially in Telecomms.
      • For under $1 million, we could come up with multiple cheap alternatives for these relatively low-bandwidth links.
      • The labour & machinery cost of fibre won't come down much over time, but communities do fund-raise over a decade or more for facilities, like a swimming pool.
      • In the meantime, do they all get an expensive 3G/4G fixed wireless service, or can we do better for them?
  • There are nearly 9,000 nodes already in the field and active. They are Telstra RIMs, a.k.a. "mini-muxes". There's also the TransACT network, covering 50,000 premises.
    • That's a long way towards the Sol Trujillo target of 20,000 nodes to cover the Big Five cities, although they aren't all in the city.
    • Neither side has talked about an acceptable short-term fix: buying or renting the RIMs from Telstra and converting them to NBN-compliant ISAM's.
    • At $20,000/RIM, is $180M for around 1 million premises. If that buys 5-7 years, it's a good deal.
      • A lucky few will get 25-50Mbps right now.
      • The rest will get good speeds until they get a scheduled GPON Fibre upgrade.
      • If the option is a standard fee for out-of-order upgrade to full-fibre, we'll soon find out people's preferences.

No comments:

Post a Comment

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