This is 'old bits', all the way back from May 96, but I'll give it a spin
anyhow. I'm looking forward to Ernie's comments, too.
You'll have to read this along with the full article.
> CRAIG MCCAW SEES AN INTERNET IN THE SKY
> Actually, he's conceived a whole new way to deliver the
> internet, using hundreds of satellites. Some think he's
> crazy. (Tell that to Bill Gates.) The cellular pioneer
> breaks a long silence to explain his vision.
> _A LONG WAY FROM TELSTAR: WHY TELEDESIC NEEDS 840
Aside from the fact we could care about this technology as comm weenies, this
is potentially the 'killer app' for a real, mass-production space industry,
not one-off government deals. And space is about as big as computers in the
Long Term... This is what Mark is in charge of, and I'm real proud of him for
that role. TD is on the buyer's side of a potential boom on the high frontier:
> Amy Sayre develops new space business for Boeing, which
> is bidding for the right to build and launch Teledesic's
> satellites. She says, "The only thing we know for sure
> is that the market is either huge or very huge. It will
> be $25 billion a year from now until 2005, excluding the
> military and Teledesic. If you count in Teledesic, the
> numbers skew."
> A few days after he sold McCaw Cellular Communications
> to AT&T, Craig McCaw went fishing with Bob Ratliffe, a
> longtime friend who headed up public relations at the
> cellular company. The moment was one of triumph, or so
> one might have thought. The AT&T deal had made McCaw
> ridiculously wealthy, for one thing. Perhaps even more
> satisfying was its vindication of all the risks he'd
> taken in pulling the cellular business out of its low-tech
> backwater and turning it into the growth engine of the
> telecom industry. McCaw had a slightly different reaction.
> "Well," he said to his friend, "I guess my career is
My kind of attitude -- pure Type A!
"His very first startup was already a vast improvement upon its successors" :-)
> has decided to create a celestial counterpart to the Internet,
Let's not overlook the massive sea change in positioning that went on here.
Back in '92-3, the glory days of ITV, these projects were the 'celestial
counterpart to the phone system' -- the bandwagon shift is critical, and a
massive improvement in their foundation. Of course, poke TD management, and
they were thinking of this all along :-)
> McCaw hopes to achieve this by creating a constellation
> of 840 satellites that will gird the planet at low
> altitude, like a rotating coat of chain mail, transmitting
> signals from any point on the planet to any other with
> the speed and capacity of fiber-optic cable. He calls
> the venture Teledesic. Others call it crazy, offering up
> the same kind of scorn McCaw received before, when he
> hocked himself to the gills to buy cellular phone
> licenses. McCaw was right then. If Teledesic succeeds,
> he can repeat his trick, and transform telecom again.
> AT&T paid $11.5 billion for McCaw Cellular in 1994; in
> 15 years, some telco may fork over many times that amount
> to buy Teledesic.
LEVERAGE. syn, cojones. No mere $10m Java play, this is for serious bux.
And if it succeeds, it may well be worth the multiple. But it will always be
a premium service -- assuming it's a $10bn capital investment for TD (current
estimate is $9B), it must throw off $1B annually in subscriber charges.
Amortizing the debt, adding operating costs, taxes, suppose this is at least a
$2b/yr business to start in the black. Sounds doable, since that's a fraction
of the world interexchange bwidth market. It was insane when it was supposed
to come from the pockets of 1m wired millionaires (there are only 26m
millionaires on the planet; apply 97% theorem).
> As with cellular, the technology behind Teledesic is not
> rocket science. Says Nobel laureate Arno Penzias, chief
> scientist at Bell Labs: "Nothing here violates the
> technology boundaries as we know them. They're not asking
> for mental telepathy or antigravity. Launching a low-orbit
> satellite has certainly been done for generations, and
> the idea of mass manufacture applied to this kind of
> technology seems perfectly straightforward."
CHUTZPAH. The only missing ingredient in this mix. But, I think like most
projects, it could be held hostage by software development. Ironically, the
real 'rocket science' is lofting mobile, agile routers in LEO. Mobile Mesh
networking (mobile routers, as opposed to mobile hosts) is only *beginning* as
a military/IETF working area. And remember, if MMnet software really works
well enough, Rajit's processors could trounce this system on the ground for
lower bwdith, higher latency applications -- the combo router/wristwatch could
build an instant internet anytime enough people are in a city!
> What makes Teledesic so breathtakingly audacious is
> McCaw's vision, the breadth of what he wants to do with
> this technology. The cellular business, by comparison,
> was simple--just a more convenient way to use an old and
> familiar service. Teledesic puts a new spin on a new set
> of services, and it inherits all the uncertainties that
> shroud the Internet: How many people will use it? What
> they will use it for? What will they be willing to pay?
BS! I think this graf is 100% backward. The audacity is lofting cellular
service. How many occasions are there to use vox capability from anywhere
remote enough to require TD service? Vox is not continuous, strike two; data
drives the need for continuous uplink. Basically, cellular application profile
is/was inconsistent with the capital structure of this project. DATA, on the
other hand has a built-in demand and, soon, inherent mobility. It scalability,
burstiness, and potentially extremely high-value content can make a business
of this. Most importantly, as far as guaging the risk of this project, there's
a lot GREATER risk in trying to appeal to the cellular metaphor in this game;
the lessons of domestic data and VSAT services may be, in fact, more
Also, the buyers are more manageable:
> Teledesic won't be much help to Himalayan trekkers who
> want to call in rescue helicopters with a cell phone.
> Rather, it will cater to computer users wishing to send
> and receive data at high speeds, especially those who
> live in remote areas. Teledesic will bring them a chunk
> of radio spectrum meaty enough to deliver a
> videoconference or a high-speed data dump. They will pay
> for the service only for as many seconds as they need
> it, and at a price comparable with a high-speed hookup
> over the regular phone network. ... Whatever the
> application, end users won't pay Teledesic directly; it
> will contract with local telephone companies that will
> market the services.
> His thoughts often seem to progress in a nonlinear
> fashion, which McCaw says stems from dyslexia. He will
> shift from one thread of a conversation to another, to
> a third, then back again, intertwining the ideas like
> strands of a rope. He has difficulty absorbing lengthy
> written documents and usually avoids them. That leaves
> time for him to do what he prefers anyway, which is to
> think and to stand back and take in the big picture.
> Dyslexics often succeed in the arts and in fields where
> spatial representation is important. Winston Churchill
> was thought to be a dyslexic. So was Albert Einstein.
> Charles Schwab and Thomas Watson Jr. are among the
> dyslexics who have excelled in business.
> Though it's a myth that dyslexics universally see things
> in reverse, McCaw gives flesh to the metaphor, saying
> that he is good at seeing circumstances from the other
> person's point of view, or at least in a different way
> from most. That helps him do what great entrepreneurs
> do, which is not to invent but to see the hidden value
> of an idea already in plain sight, a value that seems
> obvious as soon as it is given voice. McCaw didn't
> discover wireless communications--he was merely the first
> to truly understand what it was worth.
Normal people always seem so much less interesting by comparison, right?
> he is now one of AT&T's largest individual shareholders,
> he has refused to take a seat on the board because he
> can't stand going to the meetings.
> The AT&T deal made $2.3 billion for McCaw and his three
> brothers, and he has used some of his money to invest in
> a variety of startups. He doesn't need the cash these
> enterprises might bring; what he wants is a career.
> He missed his first chance to ride the Internet wave,
> for example. He recalls: "Two or three years before the
> Netscape IPO, Steve Jobs came to Seattle and we were just
> chitchatting. He said, 'The Internet is it. I think it
> is the greatest change coming in computing.' And I
> thought, 'It sounds great. How do we buy it?' Well, it
> wasn't quite that easy." He did look into buying a piece
> of one or another of the companies that connect users to
> the Internet, like UUNet, but thought they were all
> overpriced. "And they weren't," he says. He pauses. "It
> was a focus issue. We were winding down McCaw Cellular,
> and I suppose that was part of the excuse."
Steve's magic hand again... he DOES live for blood in the streets and
crusades, eh? BTW, I shudder to think that 'Netscape IPO' could define the Net
era like 'IBM PC released' did.
> It is counting on growing fiber-optic networks in the
> developed world to create and meet demand for high-speed
> services. Teledesic would fill in the blanks everywhere
This sparked a nifty realization: I was thinking back to the massive (6+ mos)
problems we had getting a 128Kb line installed from MIT to INRIA (and it
still isn't reliable). The barrier here was NOT hardware, or access to fiber,
or developed infrastructure. It was the phone company bureaucracy: FT and MCI
added those 6 months, pulling, testing and routing cable (has to go through
Paris for regulatory reasons, for example (presumably for DGSE's tapping
convenience!)). I'm not worried that INRIA isn't physically served, I care
that it's not easy-on-off deployable, like local-loop US service is (new
centrex lines in just minutes).
Translation: 'blanks' are places in the lawbooks, not just on the map.
> "AT&T can't own too much of it. We need global partners,
> because it's a global project. While it may have some
> American ownership, we can't appear to be a mere shill
> of the American industrial establishment. We've got to
> cooperate with everybody in the world."
One more sign the world is changing radically around the millenium. Slowly ut
surely, wealth IS multiplying and the G-7 WILL become the G-17 and then the
G-70. Perhaps even in our lifetimes, we will see properity extend south from
the small bands of wealth that fund our life today. This technology is a key
ingredient. Another is the amazing likelihood that Arthur C. Clarke's dream
may yet come to pass by 2001: abolishing telephone charges, to build a global
community on flat fees, uninhibited by distance in miles or cultures. I never
believed it as late as 1992 or so, but now, we are already on course to
replace the world's nervous system of expensive circuit-switched copper with
flat-fee packet switching...
> To oversee Teledesic's day-to-day affairs, McCaw has
> cherry-picked veterans who span the breadth of the
> infotech industry. Daggatt is a super-bright international
> telecom lawyer who is navigating Teledesic through the
> treacherous proceedings of international regulatory
> bodies, whose imprimatur is essential for any large-scale
> satellite project. David Patterson, the engineering
> director, helped design Sprint's fiber-optic network.
> And in a staffing coup, Teledesic has just hired as its
> chief network architect one of the designers of the
> Internet, Hans-Werner Braun.
I wish I knew a lot more about the folks at this level, to better judge the
team they've assembled. I can say one thing: this globe-girdling project is
one of the only where I can believe the lawyers are legitimately as important
as the scientists and engineers :-)
> Plans call for a first satellite launch in 2000, just
> four years from now, and the launch of all 840 satellites
> over the following two years.
Think about it: one new satellite PER DAY!
> But Teledesic has not yet settled on satellite design.
> It has not developed the software that will control the
> network. It hasn't decided how to transmit signals from
> satellite to satellite (it must choose between radio
> waves and laser beams). Nor has it lined up launch
> facilities, a major consideration given that Teledesic
> will need as many satellite launches in 2000 and 2001 as
> the rest of the world combined.
The software snafus with DIA's baggage system will pale by comparison :-0)
Actually, this is one micromarket that REALLY values good software processes.
The best sw teams in the world work on this problem; two SEI CMM level 5
teams are the US space shuttle sw maint and Motorola's competing Iridium sw
team (in Bangalore!).
> One thing he's betting on is that digital technology will
> continue its furious advance. His staff was about to
> spend half a million dollars to start writing software
> to control the satellites. He stopped them. "I mean, half
> a million dollars in software?" asks McCaw. "And spend
> five years writing it? It will be out of date before we
> get it done." McCaw will leave the job till later, when
> he can better estimate what hardware will be available
> to run the software that will run his satellites.
OTOH, the 'scary' thing is that the whole project may be circumvented by the
market and the problem reduces to lofting off-the-shelf Cisco 9000 or 10000s.
> For now, McCaw wants Teledesic to focus on what may be
> its toughest job: figuring out how to build all those
> birds and get them up without breaking the bank. (See
> box.) Following Motorola's lead with Iridium, Teledesic
> is completely rethinking the way satellites are
> manufactured, transforming the process from a craft to
> mass production. Says Jim Geros, business development
> manager for Boeing's defense and space group: "Today
> you have lots of engineers in bunny suits praying to
> one satellite. Teledesic can't do that, or they'll never
> hit their cost target." The target is $5.5 million per
> satellite--way under the $100 million it can cost to
> build one of today's big communication satellites, and
> about half what Iridium is planning to spend. Smiling
> broadly, Geros calls the goal "a big challenge. But I
> won't say it's not achievable."
> The bunny-suited guys adopt a worshipful attitude because
> a whole universe of revenues can ride on the $100 million
> object they are fashioning, and it must be almost as
> perfect as God. Teledesic, instead, will build satellites
> the way a company like Gateway makes personal
> computers--screwed together from components by many
> different manufacturers, tested quickly, and launched.
> McCaw expects that speed to beget imperfection. But if
> it works as planned, Teledesic, like the Internet, will
> be almost self-healing. When a computer relaying traffic
> on the Internet fails, or when a backhoe severs a
> fiber-optic line, the Internet instantly reroutes traffic
> over alternate routes. Teledesic's fine-meshed web of
> satellites will perform the same feat. Like a self-sealing
> tire, it will continue to roll even if a hole appears.
> And at the lower prices McCaw expects to pay for
> satellites, Teledesic can afford spares. As Daggatt says,
> "We're building reliability into the network."
So now you see why McCaw started by hiring Mark, a mfg dude, instead of
computing and comm visionaries and network engineers. I have to admit, I
hadn't thought it out yet myself. As a self-centered sw type, I figured that
would have to start very early in the cycle; I all but threatened Mark TD
would sink if certain sw routing breakthroughs made terrestrial wireless
radically cheaper, and thus sw was almost the first or second order of
> Once again, McCaw has a grander vision than anyone else
> in telecommunications. As he talks about his scheme, he
> seems to be wrestling with the social changes that might
> ensue if all the world is put within instant reach.
> McCaw's speech is laced with an idealism that seems quite
> sincere. He believes, in essence, that communications
> can help even out economic advantage.
> McCaw puts it like this: "Here's a poor village in
> Guatemala. They have solar-powered electricity; they have
> television; they see our riches and they want them. But
> where they are, they cannot have them. They don't have
> communications, and they don't have the tools to make
> money. Yet they have crops or they weave blankets, things
> that could be quite valuable if there were not so many
> middlemen, if they could essentially be a part of the
> world market."
> If that becomes possible, he argues, indigenous societies
> will be able to survive, rather than disintegrate as
> young men and women leave to seek work in the city. United
> Nations figures show the world's urban population swelling
> by 168,000 every day, and with mass migration come scary
> consequences. "Whenever you add urban infrastructure,
> you ultimately destroy everything that came before," says
> McCaw. "It's like dragging the plague around behind you.
> The beauty of electronic technology, unlike cars and
> freeways, is that we can resolve problems that are
> completely intractable when you move people physically.
> Moving electrons gives us flexibility."
Good thing we can always be certain no good deed will go unpunished by the
Left, though. Pike surfaces again, still fighting the Star Wars debate crossed
with Ralph Nader:
> As John Pike, a satellite skeptic at the nonprofit
> Federation of American Scientists, says, "I think it is
> extremely unlikely that Teledesic is going to happen or
> make any money. On the other hand, if it does happen and
> does make money, it will make a preposterous amount of
> money. It will make an amount of money the likes of which
> no one has ever seen in recorded history, because
> basically they will have established an impervious
> monopoly in whatever markets they manage to penetrate.
> They'll be in the position that AT&T was in before Judge
> Greene got ahold of them."
--- Rohit Khare -- World Wide Web Consortium -- Technical Staff w: 617/253-5884 -- f: 617/258-5999 -- h: 617/491-5030 NE43-354, MIT LCS, 545 Tech Square, Cambridge, MA 02139