A succession of waivers is no substitute for a good set of rules.

When the FCC first proposed to authorize ultra-wideband technology, back in 1998, the idea was controversial, to say the least. Ultra-wideband distributes a signal across a very wide range of spectrum, sometimes many gigahertz, but at very low levels. The maximum emissions proposed (and eventually adopted) are the same as the stray radio-frequency emissions allowed from an iPod or a digital alarm clock: at or below 75 billionths of a watt. Yet spectrum occupants of every stripe (see a partial list here) rose up as one to object. Among the opponents were the FAA and other Federal agencies, arguing that ultra-wideband signals could interfere with critical safety equipment, such as aircraft radar and communications, and thereby pose a threat to the public.

The FCC responded with a compromise: eight separate categories of ultra-wideband devices, each with its own rules. Each has a different set of maximum emissions at various frequencies. All of the categories have limitations: indoor only, hand-held only, sale only to police and fire departments, etc. All but two have limitations to a specific use: medical imaging, ground-penetrating radar, and so on. The purpose of the restrictions is to limit the spread of devices that operate at frequencies used to provide critical services.

The eight categories reflected the state of ultra-wideband technology at the time the rules were adopted, nine years ago. But the engineers have stayed busy since. Predictably, not all of the later-emerging technologies fit within the original scheme. Many of the newer devices need an FCC rule interpretation or waiver before they can be marketed in the United States. We count about a dozen waivers after the 2002 rulemaking, either granted or currently out for public comment. Not included are requests that were later withdrawn, and other requests that are pending but not yet on public notice. Also not counted are private letter rulings to individual companies, the number of which is not publicly known.

This high level of activity – more than one waiver a year – suggests the ultra-wideband rules are badly in need of revision. To its credit, the FCC expected as much. The very first paragraph of the 2002 order promised, within the ensuing six to twelve months (i.e., by April, 2003), to issue “a further rule making to explore more flexible technical standards and to address the operation of additional types” of UWB devices. And, to be sure, the FCC has since added a few more categories called “wideband” systems in a different part of the rulebook. But the original ultra-wideband categories rules remain unchanged. And the waivers keep coming.

The most recent concerns two separate categories, one covering devices that can be used only for imaging within walls, and another for devices that can be used only for imaging through walls. Needless to say, the two have different technical rules and sales restrictions. A within-wall device is further limited to use with walls that are thick enough to absorb most of the signal. Specifically excluded are “stud locators” of the kind typically used on walls made of thinner materials. A qualifying within-wall device can be sold to the construction and mining trades, among others. A through-wall device has none of the limitations listed above, but can be sold only to state and local police and firefighters.

A company called Robert Bosch wanted to sell an in-wall device to construction professionals. But its product is capable of locating studs, among other functions, and the company could not guarantee it would never be used on thin walls. Under this combination of circumstances, it satisfied neither category, and was ineligible for sale. The FCC responded with a waiver, in part on a finding that the device does not increase the potential for interference to other services.

The Robert Bosch waiver is typical of those that have gone before: a non-interfering device fails to satisfy the rules only because it falls through the cracks among the various categories. Indeed, at least one product needed a waiver because it was less interfering than a compliant device. 

Regulation by waiver has disadvantages. The process is slow and expensive. Most waivers take a year or two (at six months, the Robert Bosch has broken all recent records), and consume significant legal fees or the equivalent in other resources. The unpredictability of the outcomes hinders business planning and can make investors nervous. Worst of all, the end result is an odd patchwork of boundaries between permitted and non-permitted devices that makes no coherent sense.

Much better would be a revision of the rules that eliminates the ultra-wideband categories and instead lays out a uniform set of technical and marketing requirements. The FCC knows how to do this. For decades, the Part 15 rules for unlicensed devices were category-based, like the ultra-wideband rules are today. Every new application required a rulemaking or a waiver. But in 1989 the FCC made the switch to technically-based Part 15 rules that are largely independent of applications. There followed a surge of innovation in unlicensed devices that included Wi-Fi, Bluetooth, RFID, and countless other developments. 

The initial ultra-wideband proceeding was politically difficult for the FCC – that shows clearly in the record. Perhaps granting individual waivers is less painful than dealing with the underlying rules. But the present system impedes the development of new and potentially valuable technologies. The sooner the FCC can face up to the task of simplifying ultra-wideband regulation, the sooner the country can reap the benefits of its engineers’ creativity.