O3b Networks LEO Satellite Internet

[Danderman]

http://www.o3bnetworks.com/

In 2010, bandwidth access will get cheaper, faster and easier for most of the world. O3b Networks is building a new telecommunications infrastructure offering low-latency links from 1 Mbps to 10 Gbps for core trunking, instant fiber-path restoration and 3G Cellular backhaul.

 

[JWag]

LEO Satellite Internet, eh?

Here's their bird.  With "steerable antennas" it won't be for individuals, but ought to be able to deliver bandwidth to remote areas without having to run cable. 

I wonder who will launch them.

[braddock]

I'm not giving these guys much credence, based on their market-speak web site, simplistic looking satellite sketch, and big expensive PR splash today. 

BBC World Service was strongly playing O3b's "save the world" angle this morning. 

Of course if Google is *really* behind them then they can afford to make any fairytale they want to come true.  But I get the impression they are drawing just a little from a large Google VC fund along with HSBC.

[pippin]

But it's time for that business case.
The first tries were too early (Bill Gates, don't remember the name of the network) or too much focused at phone and direct reception (Iridium).
But it's quite obvious that
a) You DO want internet access in every ever so remote spot of the world these days, and
b) that you cannot do that with cables.

The idea is pragmatic: provide "backbone" access for local distributors. Worked for cable-TV, can work here. Google has funded worse investments.

[Jirka Dlouhy]

BG´s network was named Teledesic

[pippin]

BG´s network was named Teledesic

Thank you.

And - of course - I forgot to mention Orbcomm!

[Comga]

I'm not giving these guys much credence, based on their market-speak web site, simplistic looking satellite sketch, and big expensive PR splash today.  ....

Your other comments aside, it is hard to understand why the satellite image is so "simplistic".  Based on this news release (From Thales but posted on the O3b website) they are into fabrication.

Cannes, September 9, 2008 – Thales Alenia Space today announced it has started construction of 16 of its Low Earth Orbit communications satellites for Jersey Channel Islands based O3b Networks Limited, with options for additional spacecrafts. The satellites will offer low latency Internet backhaul to emerging markets and developing countries worldwide at speeds up to 10 Gbps with a combined total capacity in excess of 160Gbps. Thales Alenia Space started design activities under contract in 2007 and has received Authorization To Proceed (ATP) with the construction phase of the project. The delivery of the initial satellites and service activation are scheduled for late 2010.

[braddock]

I'd like to hear some technical details about how their satellite system is expected to work.
Thales calls them LEO comm sats. 
Can 16 provide constant coverage, and in what type of orbit?  Sounds more like a MEO bird to me.  Or is the plan for many many more?
Are those steerable directional dishes in the illustation? 
Are they mechanically tracking a few subscriber spots as the bird orbits? 
Or is it a steerable phased array?

I'm not skeptical of their overall technical approach because I'm sure they've been vetted.  I'm just curious to peek through the market-speak, and am uncertain of their current design stage (no Thales PR does not reassure).

I'm always skeptical of a startup that comes out of nowhere in a well orchestrated PR explosion while dropping names, and holding less than 1/10th of the amount of money they need to make a dime back.  That is like a plea for help.

They apparently have about $60 million in investment.  They estimate that they need another $640 million to pull it off.  They hope to get a loan for most of that (I can't imagine...).

http://www.bloomberg.com/apps/news?pid=20601103&sid=aqjBH48FtCbk&refer=news
 

[synchrotron]

I'd like to hear some technical details about how their satellite system is expected to work.
Thales calls them LEO comm sats. 
Can 16 provide constant coverage, and in what type of orbit?  Sounds more like a MEO bird to me.  Or is the plan for many many more?
Are those steerable directional dishes in the illustation? 
Are they mechanically tracking a few subscriber spots as the bird orbits? 
Or is it a steerable phased array?

The round-trip transmission times are also consistent with a MEO constellation.

[Danderman]

Some nuggets from their web site:

"For emergency services we can locate significant capacity within 10 minutes anywhere in the world between 45 degrees N/S. "

"O3b Networks' system virtually eliminates the delay of standard GEO satellites by reducing the round-trip transmission time from over ½ second to just 1/10 of a second. The reduced round-trip delay creates a web experience closer to terrestrial systems such as DSL or Optical Fiber."

This would work out to a MEO constellation at about 8,000 miles or so altitude, at a very low inclination.  Perhaps the plan would be for SpaceX to launch the constellation, at 9 degrees inclination. I wonder if there will have to be ITU filings for this constellation, since it would seem that for some users, these sats would be in the line of sight of GEO comsats and the ground.

[Danderman]

Since the Falcon 1, even the enhanced version, is far too small to put appreciable mass in a MEO orbit, there are two realistic scenarios for this constellation to be launched:

SeaLaunch: A single mission could probably put a quarter of the constellation in place, so if SeaLaunch prices are anywhere near $100 million, then the estimate of $650 million total capital is way off. I am assuming here that each comsat is going to have a mass of about 1,000 kg, far more than a Globalstar comsat in LEO, but far less than the standard GEO bird.

SpaceX: If a Falcon 9 has a price of about $40 million, in its normal 2 stage version, maybe 2 comsats could be launched at a time, which leads to 8 missions, at $320 million total. This is possible under the $650 million development cost.

In either case, a multiple satellite dispenser would have to be developed, and they are not cheap.

I may be overlooking an alternative vehicle that could put 1,000 kg (ie a single spacecraft) into a MEO orbit, but such a vehicle would require 16 launches, and so the risk of a launch failure would be must higher, plus each launch would have to cost less than $20 million, otherwise SpaceX would be a better option.

[aero313]

Teledesic, Skybridge, Ellipso, ICO...  I guess some folks never learn.

Lessons from the 1990s:

1)  Paper constellations begat paper rockets (which are ALWAYS less expensive)

2)  Ground antennas are ultimately less expensive than satellites

3)  The only way constellations make financial sense is if you can buy one for five cents on the dollar at the bankruptcy sale. 

Has anyone asked the original investors in Iridium, Globalstar, or Orbcomm how those investments worked out for them?

[Danderman]

Teledesic, Skybridge, Ellipso, ICO...  I guess some folks never learn.

Lessons from the 1990s:

1)  Paper constellations begat paper rockets (which are ALWAYS less expensive)

2)  Ground antennas are ultimately less expensive than satellites

I am not sure what "ground antennas" means in this context, but assuming its "terrestrial communications systems" such as cell phones, this is true to a limited extent. I think that Google and the other investors are looking at a market segment that will not be covered by terrestrial communications systems, however. What is not clear is whether it is profitable to provide such coverage to markets that will not ever enjoy cell phone or landline service.

It is a good question whether the O3b Networks people have studied the earlier failures, or whether they are plunging ahead in the overconfident-ignorant mode.

[pippin]

Teledesic, Skybridge, Ellipso, ICO...  I guess some folks never learn.

Lessons from the 1990s:

1)  Paper constellations begat paper rockets (which are ALWAYS less expensive)

2)  Ground antennas are ultimately less expensive than satellites

3)  The only way constellations make financial sense is if you can buy one for five cents on the dollar at the bankruptcy sale. 

Has anyone asked the original investors in Iridium, Globalstar, or Orbcomm how those investments worked out for them?

The world has moved on a bit since the 90s and also the business models seem to become a bit more refined.

The BIG BIG BIG mistake in the 90s was that people were trying to do direct M2S communication (mobile to satellite). #2 was they were trying to cover the whole planet.

#1 is a bad idea since it drives a LOT of requirements: you need specialized (and to be developed) technology for handsets, high power, line-of-sight connections for your handset, lots of sats and and and
O3b doesn't seem to make that mistake, they only want to provide backbone access, for which requirements are much more simple (your ground station can have line of sight and lots of power etc etc.) and - probably most important - all the technology is in place, no dev. cost for that.

#2 is a bad idea since you have to launch a lot of constellation that does not pay. Very few users in the arctics and lots of competing infrastructure in Europe, Asia and North America. You can probably get away with a lot LESS sats if you focus on your "new" Market, which seems to be South America, Africa and South Asia.

Also the need is more pressing today, you cannot run a global business without broadband internet access these days. And that includes remote office access.

I know that currently it is a PITA to deal with some areas there today. We do have local offices down there that you just cannot get connected to today's IT systems because the infrastructure is not there and that costs a LOT of money so I'm pretty sure the market is there.

I do NOT say that O3b will be successful - still a lot you can do wrong. But I am pretty sure some day somebody will be. In the end, a few sats are MUCH cheaper than thousands ok km of buried (and maintained, they get stolen these days!) cable or thousands of point-to-point comm links.

[nomadd22]

"

This would work out to a MEO constellation at about 8,000 miles or so altitude, at a very low inclination.  Perhaps the plan would be for SpaceX to launch the constellation, at 9 degrees inclination. I wonder if there will have to be ITU filings for this constellation, since it would seem that for some users, these sats would be in the line of sight of GEO comsats and the ground.

More like 4,000 miles. Round trip means up and down twice.

[A_M_Swallow]

#2 is a bad idea since you have to launch a lot of constellation that does not pay. Very few users in the arctics and lots of competing infrastructure in Europe, Asia and North America. You can probably get away with a lot LESS sats if you focus on your "new" Market, which seems to be South America, Africa and South Asia.

Ships in the middle of the sea also need voice communications.

[pippin]

#2 is a bad idea since you have to launch a lot of constellation that does not pay. Very few users in the arctics and lots of competing infrastructure in Europe, Asia and North America. You can probably get away with a lot LESS sats if you focus on your "new" Market, which seems to be South America, Africa and South Asia.

Ships in the middle of the sea also need voice communications.

This is about broadband data.
Plenty of solutions for voice available today.
Few ships per square mile.
But ships will also need broadband data.
But, again, then most ships are traveling at moderate latitudes, have no line of sight limitations and plenty of power, so it should just work fine.

[anik]

Sea Launch Signs Launch Agreement with O3b Networks
http://www.sea-launch.com/news_releases/nr_082309.html
 
Long Beach, Calif., Sea Launch Company has signed a launch services agreement with O3b Networks Limited for up to two launches in support of their Medium Earth Orbit (MEO) telecommunications satellite constellation, with the first launch set for late 2010.

Built by Thales Alenia Space, the O3b Networks satellites will be deployed by Sea Launch's Zenit-3SL system in groups of eight per launch, to an equatorial injection orbit of 7,825 kilometers above the Earth. The satellites, each weighing approximately 700 kg (1,540 lb.), are designed to provide high-speed, ultra-low-latency Internet Protocol (IP) connectivity between emerging and developed markets worldwide.

O3b Networks' new global communications backbone will fundamentally change the way IP, 3G Cellular and WiMAX networks interconnect. This will reduce the costs of backhaul for Mobile Operators and Internet Service Providers, enabling them to provide more cost-effective services in underserved and emerging markets.

Sea Launch is developing a new multi-spacecraft dispenser for these missions, which will accommodate O3b Networks’ specific orbital insertion requirements. Upon successful sequential deployment, the satellites are expected to have an on-orbit maneuvering lifetime of ten years.

"We selected Sea Launch because of its knowledge, expertise and successful track record of moving payloads into orbit," said Greg Wyler, O3b Networks Founder and CEO. "With the placement of our first eight satellites, we will provide emerging-market network operators with a low-cost, high-speed alternative to connect their 3G, WiMAX and fixed-line networks to the rest of the world. This will allow consumers and businesses in emerging markets to benefit from high-speed Internet connectivity for educational, medical and commercial applications."

"We are delighted to have the opportunity to launch these innovative spacecraft for O3b Networks and to have a significant role in this dynamic new communications system," said Kjell Karlsen, President and General Manager of Sea Launch. "We are grateful for O3b Networks' confidence in our reliable launch system and the team behind it. We are proud of our ability to not only launch satellites, but to help launch O3b Networks’ new communications system, which will bridge the digital divide substantially."

[Danderman]

OK, so I was 125 km and 300 kg off in my estimates for this constellation. Fortunately, this allows O3B Networks to launch their entire constellation in 2 flights.  This also means that SeaLaunch must develop an 8 satellite dispenser, which I suspect will be closely related to the 12 sat Globalstar dispenser.

I find it interesting that the mass of each comsat is 700 kg, which is very low mass for a direct broadcast satellite; Globalstar satellites have a mass of 400 kg, but they provide much less bandwidth, at only 1400 km altitude. I guess transponder technology is much improved since Globalstar.

The last point may be more critical in the long term: how much more massive would a small SPS satellite have to be in this orbit? Assuming that one Falcon 9 could put an SPS in an 8000 km MEO orbit, perhaps we may see an equatorial SPS global system sooner than expected (albeit at a low power rate).

[pippin]

I find it interesting that the mass of each comsat is 700 kg, which is very low mass for a direct broadcast satellite;

If I understand their model correctly, they're nor direct broadcasting sats.
That's what they do better then Globalstar.