Take a Number

By eweek  |  Posted 2001-05-07

Take a Number

Fantasma networks boasted some of the best minds in the world of ultrawideband technology. Backed in part by Microsoft co-founder Paul Allen, the company dreamed of creating a new kind of wireless data network on the cutting edge.

Last month, Fantasma closed its doors. The reason: The company could raise no more money because the Federal Communications Commission has yet to set any technical standards for ultrawideband transmissions. And the FCC is a long way from doing that, because its small engineering arm is plagued by chronic understaffing, aging test and measurement equipment, and anemic budgets.

Those problems are costing technology companies untold millions of dollars--in some cases sucking the profit entirely from products with short shelf lives. The typical mobile phone, for example, has a market life span of a little more than one year--with most of the sales coming in the first six months. Any delay, even a few weeks, can be very expensive.

"The cost is in the millions and millions of dollars," said Grant Seiffert, vice president of government relations at the Telecommunications Industry Association.

If the problems persist, companies fear they will also interfere with the deployment of promising new technologies such as Bluetooth, a wireless standard for automatically interconnecting mobile devices in close proximity to one another. And that could threaten the already precarious financial situation of other manufacturers.

"Bluetooth is becoming more or less a standard in products coming to the FCC," said Allen Groh, director of international trade and regulatory compliance at Ericsson. "And it takes a very long time to get a new engineer [who understands it] online and train them."

Old System, New Technology

Old System, New Technology

From ultrawideband to bluetooth to wireless data phones, PCs and screw-in fluorescent light bulbs, the FCC is a vital stop on the road to market. No product that emits radio waves--either as part of its function or as a byproduct--can legally be marketed in the U.S. without meeting the FCCs requirements And for most mobile products and leading-edge technologies, marketing requires advance FCC approval that depends on a review of test data by 14 engineers in suburban Maryland.

The newest technologies have to wait while the skeleton staff--working with old equipment and ever-changing political pressures--devises the tests and specifications the products must meet to be approved. At the same time--and often at the request of competing companies--they must take time out to test samples of products already certified and on the market to make sure the real McCoy is behaving the way the paperwork said it would.

To top it off, the engineers must also produce one or two major reports per year "to determine the interference impact of changes in spectrum sharing provisions, or reports on the impact of new technologies such as digital television," an FCC spokeswoman said. And that is one area in which the outdated equipment makes for a daunting task.

All that adds up to a huge backlog at the FCCs labs. While most agree there has been a significant reduction in the past two years, hundreds of products are still gathering dust, awaiting approval. They belong to some of the leading names in high-tech: Alcatel, Ericsson, Lucent Technologies, Motorola, Nokia and Seimens, to name a few.

"Time to market is very critical to this industry," Groh said. "A company can lose as much as $1 million a day if they are trying to introduce a device with new features that cant go on the market for 60 days."

Even the FCCs chairman, Michael Powell, has told Congress his laboratories are outdated and the agency is in danger of losing some 40 percent of its engineering talent within the next four years, as the engineers reach retirement age.

"The lab itself is in growing disrepair," Powell testified on March 29. "We dont even have the equipment that will even measure the [products]."

But the FCC has received little sympathy--or money--from Capitol Hill recently. It was one of three agencies targeted for elimination after the ascent of former Rep. Newt Gingrich, R-Ga., and the Republican House in 1995. It has since been caught in a tug of war between Democrats who support the agency and majority Republicans constantly looking to privatize or eliminate its functions.

"Generally speaking, unless you have support from people over on the Hill that sort of believe in what youre doing, and people from OMB [the White House Office of Management and Budget] that believe in what you are doing, its hard for a regulatory agency in this deregulatory climate to make the case for funding," said Dale Hatfield, a longtime FCC engineer who served as head of the labs from 1997 until December 2000.

Companies Pay the Price

Companies Pay the Price

While that lack of support has cost the agency and its labs funding, technology companies--and their customers--are the ones that ultimately pay.

"The combined risks were too high to raise money in this capital market," said Roberto Aiello, Fantasmas founder and chief technology officer. "And the regulatory uncertainty was the highest perceived risk by the potential investors."

Fantasmas case is unusual. It faced not only a bear market, but also claims by other powerful interests--among them the Federal Aviation Administration, the Department of Defense and the global positioning system industry--that ultrawideband would interfere with air traffic radar or GPS receivers. It has been more than two years since the FCC began to look at those claims. But, in part because of the lack of equipment and expertise at the labs, the FCC is in no position to resolve the interference arguments before the end of the year.

In comparison, "regulatory approval in Europe may be done by the second quarter next year, and they just started on the process three months ago," said Jeff Ross, director of corporate development at Time Domain, the company that holds many of the key patents for ultrawideband.

"When a new technology comes out, companies dont know how to make it well and the labs dont know how to test," said Kenneth Nichols, chief of the FCCs laboratory division. "Those devices are most likely to have problems."

Yet it is not just new technologies that suffer. Even routine applications for approval of mobile telephones can take weeks, as the harried engineering staff plows through backlogs of paperwork. It can take six to eight weeks for a new model to be approved--in a market where the average product life is 12 to 18 months and most of the profit is made in the first months of sales.

By contrast, the Europeans last year adopted the Radio and Telecommunications Terminal Equipment Directive, which allows for an industry self-certification plan for most telecommunications and mobile devices. Manufacturers can issue a Suppliers Declaration of Conformity (SDOC) attesting that the new product meets uniform European standards, and then bring the product to market in all 15 member countries of the European Union without government approval.

Instead of premarket testing, the European Commission relies on market surveillance to catch companies whose products dont meet the specs.

A form of the SDOC process is allowed in the U.S. for some products where the technology is mature, such as PCs and digital devices that dont include transmitters. All a company needs to do is bring the device to an FCC-approved private testing lab, which can certify compliance with the rules. The SDOC process also applies to devices that are directly plugged into the public switched phone network.

But few mobile devices destined for the U.S. market can be sold using self-certification. Some may be certified by one of the 15 or so private Telecommunication Certification Bodies (TCBs) approved by the FCC, but many must be directly approved by the FCC--either by engineers examining and accepting test data provided by the manufacturer or, in a handful of cases, by FCC engineers conducting the tests. That, coupled with a shortage of engineers, has led to a big backlog of applications.

Industry executives such as the TIAs Seiffert claim there are some 650 products waiting for approval. That figure is disputed by the FCCs Nichols. Last June, Nichols said, the labs had about 700 applications pending, "which the industry felt was unacceptable." But a deliberate program to clear the backlog has slashed that number by more than half, he said.

Still, at the end of April, the FCC said there were 271 products waiting for approval, including 126 applications filed during the month.

On average, it takes 42 days to approve the wireless and cell phone products. Thats 12 days longer than the 30-day goal, and just a few days shy of when "we start getting lots of complaints from companies," said Bruce A. Franca, acting chief of the FCCs Office of Engineering and Technology.

And those figures are good only when the engineers dont have questions about the test results the companies have filed for review. A single question--often asked because a company has failed to provide all the needed information in the correct way--might add a week or two to the process; several questions could add months.

"We have tracked the process, and the actual time for certification is directly proportional to the number of questions that are asked," Ericcsons Groh said. "We try to produce perfect applications, to minimize the number of questions."

"Theres a tension here," Franca said. "We have to do our job, but the applications have to come in and the checks have to clear, so [the engineer] doesnt see the application for a week or so. Our goal is to meet that 30-day timeframe."

Stretching the Limits

Stretching the Limits

For one particular type of test, which measures the rate of absorption by the human body of radiation from things such as mobile phones, the FCC has just a single engineer capable of doing the work. While both his colleagues and industry praise his skills, the volume of gear entering the market is overwhelming.

Still, he is one of the lucky ones. The equipment he uses for so-called Specific Absorption Rate tests is one of the few state-of-the-art setups at the FCCs suburban Maryland laboratories, where most of the best gear is only slightly younger than the 1970s-era building it occupies. The SAR setup, a $200,000 installation with a precision industrial robotic arm, highly sensitive antenna, computers and a hollow "body" cavity to mimic humans, was purchased on a one-time budget grant.

All that is for testing something that has no definite risk. Neither the FCC engineers nor other scientific studies have proven any link between the radiation and medical problems. But the political side of the agency made the test a requirement based on a 1995 Environmental Protection Agency report.

Compare that to the mobile van the FCC labs use to test digital TV receivers, including PC cards, by driving throughout the Washington, D.C., and Baltimore area. The equipment in the van was cobbled together with castoffs from other parts of the labs, and installed into a rusty, 1985-vintage cargo van purloined from the agencys enforcement division. Including the one cutting-edge measurement system purchased specifically for the van, the total cost was about $75,000. Thats less than 10 percent of what a TV network might spend on the same type of vehicle and equipment.

"Ninety percent of our equipment is 10 years old or older," Nichols said. "The biggest problem is that they dont have the new functions in them."

Inside the main lab building, the most advanced measurement equipment owned by the FCC dates back to the early 1980s. Much of it is top-quality Hewlett Packard gear--at least it was, at that time. Those instruments were built before the invention of Code Division Multiple Access, one of the most widely used digital transmission standards for mobile phones. They were also built before transmission on frequencies above 2 gigahertz or so were anticipated. As a result, the lab has to rely on borrowed gear or outside tests.

In fact, the newest piece of test equipment on the site, an oddly shaped, elongated aluminum box designed to permit highly calibrated testing indoors, is being borrowed from the private sector. While it has the potential to speed up testing, because it isnt subject to the weather delays of the calibrated outdoor test field, it could be taken away from the labs at any time. And the FCC doesnt have the $150,000 to build or buy its own.

"All these new devices [on the market] require new equipment to do measurements," Franca said. "Were seeing equipment going into higher and higher frequencies. We dont have all the capability to measure at all those high frequencies." And the few machines the lab does own that can measure at frequencies above 2 GHz are difficult to calibrate--again, because of a lack of proper equipment.

Hope on the Horizon


Hope on the Horizon?

The FCCs laboratory division is the largest division of the FCCs Office of Engineering and Technology. It has 14 engineers and a dozen other workers in its three branches.

The largest branch is the Equipment Authorization branch, run by Richard Fabina, which processes all applications, and identifies "new or novel products" in the market and figures out how to best test them. Raymond A. LaForge heads the Measurements and Calibration Branch, which is in charge of developing test procedures and certifying outside groups that companies can hire to test products. David L. Means runs the third and smallest branch, the Technical Research Branch, which oversees the testing of new telecommunications technologies.

Franca said he needs at least $2 million right away--$1 million for an enlarged test facility and $1 million to bring his equipment up to modern standards--plus a $200,000 annual budget for routine upgrades to keep pace with developments.

But he also needs more people. The tiny engineering staff is composed mainly of people who have been with the agency for decades.

"It would be good for [Powell] to come out on a normal workday and see how little testing were doing because of staffing," Fabina said.

The workers are universally acclaimed for being intelligent and hardworking, good solid engineers and public servants. "Those guys work like galley slaves there," Time Domains Ross said.

But they are also, for the most part, within a few years of retirement. Some are set to retire in just one or two years. One engineer, Errol Chang, passed away at the end of April.

"Were down 10 people since 1997, and this [testing and approval] process has gotten a lot more complicated," Franca said.

Replacing staff is a daunting task, in large part because of salary. The FCC, operating on the governments civil service pay scale, can at best hope to offer an engineer coming out of college about $31,000 per year. The agency is trying to find a way to move engineers into a higher civil service ranking that would pay $47,000. But doing that is complicated because it might cause other workers in the lower scale to also demand upgrades.

Even at $47,000, the FCC wouldnt be close to competitive. In the private sector, "six-figure salaries" are not uncommon for people in the same job role, Seiffert at the TIA said, especially for those with training in the areas the FCC needs most.

Hatfield said he has been sounding warnings about the engineering shortage and pay scale problem for years, but to deaf ears. Its a far cry from the 1970s, when he first came to the FCC, and each of the then seven commissioners had--by law--an engineer on staff. "The decline in the number of engineers is sad, and its one of the critical issues" for the FCC, Hatfield said.

Powell has discussed possible ways to encourage young engineers to choose the FCC over the private sector. Among them would be some kind of student subsidy program for college students studying engineering, who would in turn agree to spend a certain number of years working for the FCC.

Hatfield believes some of the billions of dollars the FCC is raising for the U.S. Treasury through spectrum auctions ought to be "reinvested" in the labs for equipment and engineering talent.

But many technology industry executives say money is not the only solution to the problem. They want to see the FCC continue down the road to privatization of the testing and certification process, as is the case in Europe. Specifically, they want the FCC to move more quickly to expand the number of approved outside certification labs and the list of products that can be certified outside the FCC itself. Eventually, they would like a system that is entirely one of self-certification with post-market enforcement.

They also want the FCC to move more quickly to implement a mutual recognition agreement between the U.S. and the EU that would let products approved in one market be sold in the other.

The idea of shifting to enforcement and away from the mountain of approval paperwork sits well with the labs leaders as well.

"Wed like to change the labs to a function of oversight," Franca said. "Wed like to get out of the approval business in a big way. Instead, we would watch the TCBs and watch the newest technology, where there arent set procedures yet, and get involved in the technical studies area, to support the commissioners in policy decisions, the technology implications of policy decisions."

Indeed, both Powell and House Energy and Commerce Committee Chairman Billy Tauzin, R-La., have said they want to move the FCC into a greater role in enforcement and less in direct regulation.

And there may be hope for the labs in the fiscal 2002 budget submitted by President George W. Bush. It calls for a healthy spending boost for the FCC, including money for equipment and several new engineers. Although the final budget numbers are a long way off, there is a renewed sense in Congress that something has to be done about a situation that has a serious impact on the technology industry and the economy.

"That was one of the first issues that we talked about when I became chairman," said Rep. Fred Upton, R-Mich., chairman of the House subcommittee on Telecommunications and the Internet. "They do need help."

Rocket Fuel