Can Bill Gates broadband satellite bet beat re-entry burnout?
Teledesic and ICO strive to lose the money hole tag
Analysis Thanks to more than a thousand pages of regulatory filing with the SEC, we now know the business plan for broadband satellite ventures Teledesic and New ICO. The companies are to be merged later this year as a result of a rescue bid by Craig McCaw London-based ICO Global Communications. McCaw injected $1.2 billion into the bankrupt company.
Bill Gates personally, rather than Microsoft, is a significant investor in Teledesic, but McCaw has some 64 per cent of the voting power of the to-be-merged companies. His wealth came from the sale of McCaw Cellular to AT&T for $11.5 billion in 1994.
Red ink spatters the filings: to May this year ICO lost $593 million, while Teledesic lost $393 million to the end of June. It is planned that the merged company will be listed on NASDAQ by the end of March next year. Considerable sums, called impairment losses, have been written off by Teledesic - $274 million in 1999 for example - as a result of Motorola's work not having the asset value assigned in the books. There is also evidence of internal dissent at Teledesic, since substantial amounts have been paid out for settlements with staff - a quarter of the staff left in April. Teledesic says it expects to be competitive with DSL on price, but it will be less competitive with fibre because of the need for a mechanically-steered antenna. Its broadband service should be suitable for remote parts of the world and rural areas.
Which is which?
It hasn't previously been entirely clear what each company would offer, although at early planning stage Teledesic pitched more for data, while ICO (which was spun out of Inmarsat) aimed for a voice and data combination. The merged company will still go ahead with two satellite systems, but Teledesic will be more upmarket. Teledesic should be able to profit from the distribution channels and customer base that New ICO sets up, as this is expected to be operational about two years earlier. An advantage could be that Teledesic potentially provides an upwards migration path - at the same time freeing up more capacity in the New ICO system.
It is estimated that New ICO (it will be renamed after the merger) will need at least $2.8 billion before it can begin a commercial service. It plans to have 12 satellites (including two spare) in medium-Earth orbit and to start service in mid-2003, with the ability to support 14 million end-user devices. The old ICO network was circuit-switched, narrowband voice and data-to-handheld satellite phones, but New ICO expects to offer a packet-based system, with the considerable advantage that users would be able to use the system indoors.
Contractors as roadkill
Teledesic will need substantially more investment than New ICO, but the amount will not be known until a new prime contractor has been chosen early next year. Teledesic has been through quite a few prime contractors: Motorola was terminated in 1998, with Teledesic having to pay a penalty for this. Previously Boeing had been the prime contractor and had also to be compensated for having its agreement terminated. Hughes received a pay-off after its stint as prime contractor. The design of the system is currently being re-evaluated - again - and it is likely that there will be a further reduction in the number of satellites. Originally Teledesic planned 840 satellites plus 80 spares in 21 orbits 435 miles above the Earth, but this proved too difficult to achieve, so was subsequently reduced to 288 plus spares. Teledesic says it expects its delay to be 70 milliseconds each way - about 3.5 times as fast as with geostationary satellites - so users would scarcely notice the delay. Commercial service is still planned for 2005.
Both companies have experienced failures in launching satellites, and without high reliability in launching, the service quality could drop if launches can't keep up with losses once the service has started. New ICO's satellites will be launched with rockets built in the Ukraine and Russia, with some being launched from Kazakhstan, which perhaps introduces some political risks. Teledesic has an agreement for four launches between July 2003 and December 2006, with provision for a year's delay without penalty.
The regulatory hurdles are mostly solved, but there could be difficulties with US law if foreign ownership exceeds 25 per cent. The ITU meanwhile requires that Teledesic offers service by 26 September 2004, or the company would lose its priority and might have to compete in a new application for its frequency allocations. Teledesic's launch plans show there is very little leeway if the ITU requirement is to be met.
The competitors are the increasing terrestrial networks, which are being deployed more quickly in urban areas where most customers are situated, as well as rival satellite systems, with the line-up so far including Hughes/Spaceway, Astrolink, Loral/Cyberstar and Skybridge. Teledesic says it expects to be more competitive on price by a factor of two to three. DirecPC, backed by Hughes, is already operational. Later this year, the Gilat2Home service is expected to offer a two-way broadband service in the US, although it will need quite a big elliptical dish. There is some potential conflict in that Gates is a major shareholder in Teledesic in a personal capacity, but Microsoft itself is a backer of Gilat, along with EchoStar.
In view of Iridium's failure - it only managed to attract 55,000 customers for the 88-satellite system that is about to be allowed to burn-up on re-entry - there can be no certainty of success for future systems, because of the technical problems and risks. Nor could there be any quick financial returns, although the major investors have very deep pockets. The good news is that is now looks as though there may be some hope for competitive satellite services for Internet access at reasonable prices in a few years time. ®
Sponsored: DevOps and continuous delivery