Business failures are part of the reality of technology-driven visionaries.
Sometimes, however, a project is so large and audacious that its backers misread market realities despite achieving astonishing technical success. Such was the case with the Iridium worldwide satellite-based communications system.
It was a bold plan conceived in late 1987 by Bary Bertiger, Raymond J. Leopold, and Ken Peterson. They were engineers associated with Motorola, which held many Iridium patents. Although it is hard to put an exact price on the project, most estimates place the cost at between $5 -$10 billion.
At the time, Motorola was among the leaders in wireless and wireless systems. It provided basic voice radios, cellular infrastructure, and advanced systems to police, armed forces, maritime users, among others. The company was a well-known, well-respected, and well-connected supplier in more than 100 countries, with a reputation built over many decades. Without its financial investment and worldwide "influence," Iridium might never have happened.
What Iridium Is
Cellular connectivity is so ubiquitous that many people do not realize that its very availability is made possible by a cellular "backbone" network invisible to the user. This network includes linked base stations and associated mobile telephone switching offices (MTSO) located as closely as needed to provide the cell coverage (a cell can range from a few city blocks to tens of miles, depending on locale).
These MTSOs are linked not only to the base stations they directly serve, but also to each other and higher-level system management. This is usually done via fiber-optic links or point-to-point microwave links, and manages and hands-off calls as the phone user moves from cell to cell.
The problem is that most of the Earth's surface is covered with water and thus has no base stations or MTSOs; further, the maximum size of a base station is about 10 miles across (and many are far smaller). There are also large, barely inhabited areas (the Arctic poles and deserts, for example) where base stations and MTSOs are neither technically practical nor financially viable.
One alternative to a cellular stem is conventional radio. But radio offers erratic connections due to atmospheric propagation issues, interference, electrical noise, and channel crowding that can result from too many users.
Another possibility is a link to a geosynchronous satellite. But this presents problems due to round-trip latency of about 250 msec, as well as satellite visibility (which decreases as one nears either of the Earth's poles).
For ships and oil rigs, basic ship-to-shore radio exists. Even a specialized satellite system (Inmarsat) is an option, but these are used primarily for low-rate weather updates, bulletins, or critical emergency messages, not ongoing, real-time conversations.
The Iridium system, therefore, was designed to overcome all of the weaknesses of conventional radio, geosynchronous satellite links, geographical location, and travel. The idea was to set up a system which would allow cellular-like connectivity (voice and data) from anywhere on the globe.
Making this happen required a number of elements.
First, a global constellation of 66 satellites (plus about half a dozen or more "spares") would be needed. But not in the usual geosynchronous orbits or even low-altitude, low- propagation-delay orbits. Instead, polar orbits would be used for 100% Earth coverage.
The original calculations showed that 77 satellites would be needed (plus spares). Thus, the system was named after element 77 on the periodic table, Iridium. (Subsequent analysis showed that just 66 satellites were needed; re-naming the project Dysprosium for the 66th element, however, was rejected).The satellites were arranged in six orbital planes of 11 satellites each and inclined at 86.4°, just a few degrees off a perfect north-south axis; these orbital planes cross at each pole.
At a height of 485 miles (781 km), each satellite orbits every 100 minutes, with a round-trip Earth-satellite link propagation delay of 5 msec, acceptable for two-way speech.
These satellites were not just simple connecting points for one-hop, Earth-based connections between users. Instead, they were orbiting MTSOs which could dynamically link and unlink via the Ka microwave band (26.5 to 40 GHz) to any of four adjacent satellites: the ones immediately before and after in the same orbital plane, as well as the closest ones in adjacent planes.
A call in progress might go through one, two, or more satellites, and be handed off as the satellites orbited. These handoffs would occur about every 60 seconds for a given call connection. As a result, a call's link delay was the 5 msec plus the satellite-to-satellite link delay of another few milliseconds, far less than the geosynchronous link with its 240-msec hop delay.
The fabric topology and software needed to make this work were unlike anything yet orbited. The processing and routing demands were stringent, yet available power was limited to the solar panels.
Early designs measured about 4 meters by 1.8 meters, weighed 689 kilograms (1,519 lb) with innovative antennas for communicating with ground stations as well as adjacent satellites. Due to the low orbits, the satellites were in Earth-circling radiation belts and needed radiation-hardened components, yet still would have reduced life.
Second, satellites alone were only part of the overall plan. Earth-based stations called gateways were needed to link to the 66 satellites, and to each other, to complete the system management and fabric and also link to the terrestrial phone network. These gateways (there were roughly two dozen), would be located around the globe in countries which had a technology base as well as some that had nearly none.
The software needed to link these gateways and manage them in real time, along with the actual links between them, was another challenge. Communication between satellites and earth stations used 30 GHz for the uplink and 20 GHz for the downlink. Some gateways were owned by the hosting country, while others were owned and operated by a private licensee in that country, with official approval of that government. Overall system management was done by a dedicated Satellite Network Operating Center (SNOC) facility—also a major undertaking.
Third, handsets had to be provided so users could actually contact the satellite constellation. Motorola contracted with several vendors, including Kyocera and their own handset group, to design and build these. The first units were relatively large and bulky (3 pounds), and were priced at several thousand dollars. Subsequent units have shrunk considerably in size, weight, and cost, but not quite to the parameters of concurrent, mass-market standard cell phones.
Making It Happen
The technical challenges across all three aspects – satellites, gateways, and handsets – were enormous.
The design teams had to devise new topologies for these unique mesh networks, new algorithms for call handling and handoff, and new RF designs including antennas. They then had to make it all fit into a modest-sized package.
Unlike conventional communications satellites where each is uniquely designed, built, and launched, the large number of Iridium satellites mandated a "mass production" approach. The idea was to build identical satellites in clusters of 10 or more, and incorporate any changes in subsequent batches only if there was a demonstrable benefit. Thus, Iridium standardized the satellites to the extent possible, which was a novel approach.
The launches were to be be done from sites across the globe, including China, French Guyana, and the Soviet Union (before its breakup), using contracted services. A typical rocket would include at least six satellites, and sometimes up to a dozen, to save time and cost. While this provided many operational advantages, it also added risk, such as when a rocket explodes or fails to orbit and the entire multi-satellite payload is lost (which happened several times).
A separate aspect of getting Iridium off the ground had little to do with technology and everything to do with spectrum assignment, regulatory issues, and global reach.
The Motorola-driven Iridium system had to get some specific spectrum allocations from the International Telecommunication Union (ITU – the United Nations specialized agency for information and communication technologies), and also license agreements with the countries that would host the gateways. Both allocation and agreements became time-consuming bureaucratic battles, as countries objected for security, financial, and political reasons.
What eventually solved both the spectrum and gateway problems was a combination of Motorola's long-standing relationships with key people in many countries, as well as promises and guarantees to gateway operators and their governments. As a respected vendor, Motorola's local agents were able to call on government ministries and officials, get their attention, and work out needed agreements.
Launch, Crash, and Nearly Burn
The Iridium communications service was initiated on Nov. 1, 1998 with a call by then-Vice President Al Gore. After several months working through technical issues, the system worked fairly well and met most objectives. Potential users were said to be impressed. What's more, the initially skeptical technology community was impressed that Iridium went from an idea to a fully operational system in a little over a decade.
On the business side, though, things were dismal. Instead of tens of thousands of users willing to pay thousands of dollars for a phone and about $10/minute for a call, the system signed up a few hundred.
Among the reasons is that the anticipated market of worldwide business travelers never materialized. For one thing, the cost and size of the handsets was a barrier. So was the fact that most travelers were in areas that offered cellular service. What's more, a technical weakness that the caller had to be outside to "see" the satellite. Conventional cell calls, meanwhile, could be completed from inside most buildings.
In less than a year, Iridium Satellite LLC (the official name) filed for Chapter 11 bankruptcy reorganization in the U.S. What took a little over a decade to create and implement took just nine months to figuratively crash. It seemed to some that its literal crash -- and burn -- come soon.
Motorola would not tolerate the $200 million/month ongoing operating costs with no prospect of eventual turnaround. At the same time, gateway owners demanded make-good on their expected revenue. And the system needed planning for additional satellite launches as spares.
Motorola said the solution was clear: write off the loss, kill the system and shut it down. Doing so included de-orbiting the satellites by directing them into lower orbits where they would burn up. The picture was bleak and apparently hopeless.
Over the next few months, Motorola issued several "firm and final deadlines" saying that if the system did not find a buyer by a specified date, the satellites would be instructed to de-orbit within 48 hours of that deadline.
Each time, the deadline was put off due to pleading from Iridium proponents, who sought new investment partners in the $20 to $100 million range around the world, including African countries that wanted both the connectivity access and prestige, as well as their portion of the ongoing gateway revenue.
Backing Away from the Brink
A break came for an unlikely place when a New York Times article mentioned that some of the satellites might not burn up completely, and that some debris might land on inhabited areas or even hit someone.
Suddenly, legal liability became the dominant issue preventing issuing the de-orbit command. This potential debris became the delaying element and bought time for more tries at survival.
Eventually, the remaining Iridium team, led by Dan Colussy, a retired CEO of air carrier Pan Am, heard about Motorola’s plans and decided he would buy Iridium. With some effort, he pulled together financing from private investors and the U.S. Department of Defense (DoD), which became convinced that Iridium was a valuable and irreplaceable resource for the military. (Colussy, has been chairman of Gemini Capital, LLC since 2009. The venture capital group invests in new technologies in aerospace, aviation, power generation and environmental industries.)
Critical to this decision was the reality that the Colussy's Iridium team – mostly aerospace industry veterans and former DoD insiders – knew the right people to talk to.
The assets Iridium Satellite LLC were purchased for $35 million (compared to a development cost of in the billions of dollars) by the newly created owner, Iridium Communications Inc., a publicly traded company. All of the gateways were shut except for a commercial one based in Tempe, Ariz., and a DoD gateway in Hawaii.
Today the company has 782,000 subscribers, with 72,000 of them from the government. The service has expanded from basic voice to broadband for higher-speed data. Developers are working on a set of satellites for even higher-speed performance and capacity, at a cost of around $2 billion. Among the commercial users are oil rigs, ships, desert mining sites, and even the year-round scientific bases in Antarctica.
The technical success, business failure, and rebirth of Iridium provides multiple lessons for advanced-technology projects. Among them:
•Technically audacious ideas and dreams can drive astounding engineering feats—with enough funding.
•A relentless champion is needed to drive the mission from the start, and revive it after setbacks.
•It can be easy to believe your own market size projections and skip the reality checks.
•Satellite production, launch, and management can be transformed into a mass-production process, in contrast to the one-off, fully customized satellites designs and associated operation.
•Rocket launches can be successfully done on a high-volume basis by using multiple contract launch services. Each Iridium launch "kit" included the multiple satellites plus fuel for their steering thrusters, battery sets, the set-up and test instrumentation, an office, and everything else needed except the launch rocket itself. Total kit weight per launch was about 500,000 pounds, with shipping arranged though FedEx, in many cases.
•Regulatory issues such as spectrum allocation and approval can be time-consuming in just one country, and can become exponentially more complex when multiple countries or jurisdictions are involved.
•Local connections and personal contacts with key influencers in each country's bureaucracy are essential, and some quasi-ethical steps are often part of the scenario.
•Small, unplanned items can have unintended consequences, for both good and bad. For Iridium, the originally non-existent concerns about unburned debris liability became a sticking point and stalled the final order for the physical demise of the satellites. That proved sufficient to allow a rescue plan to come together.
Today, Iridium is a viable mobile, satellite-based system with users who have no practical alternatives. As with standard cellular systems, its users have little or no idea what it took in terms of money, innovation, radical thinking, technology, undaunted determination, errors, and still more money to make it possible.
1. "Eccentric Orbits: The Iridium Story," John Bloom, Atlantic Monthly Press, 2016.
2. "Satellite-based telephones punch past reality's impediment," EDN, Sept. 24, 1998.
3. "Iridium Will Host Science Payloads," IEEE Spectrum, Nov. 30, 2009
4. "The Collision of Iridium 33 and Cosmos 2251,"Aerospace, December 10, 2015
6. "LEO Communications Satellites: The IRIDIUM Constellation," DeAnn Redlin, December 14, 2000
7. "The Rise and Fall and Rise of Iridium," Air & Space Magazine, September 2004