Current attempt
Powerline networking has short-circuited before. Will this time be any different?
Erik Sherman, Contributing Editor

Oh, if only things were so easy. This is a technology plagued by poor performance, few real products, and vendors that have disappointed the industry and the public time and again with broken promises. Nevertheless, optimists see success around the corner yet again.

Powerline manufacturers all say they're about to unveil new technology that will—cross their hearts and hope to die—be fast and reliable, providing dazzling new venues for broadband and telecommunication services. All the companies have to do is deliver like they've never done before, attract the right partners, and find their place among well-established competition.

Powerline networking should be a natural choice for the public, because it has clear advantages over competing technologies. Typical homes have only a small number of phone jacks, which limits the potential of phoneline-networking products. But power outlets are omnipresent, giving consumers flexibility.

"It is the only technology that is truly ubiquitous," says Jarek Chylinski, vice president of global marketing at Enikia, a vendor of powerline-networking technology. "If you go to Europe, very rarely do you find more than one phone jack in an apartment or home." In Africa and Asia, many residences lack even that one.

Jarek Chylinski

Powerline networks to date have suffered from reliability problems, with equipment often failing to work even in otherwise functional outlets. However, the question of bandwidth—more accurately the lack thereof—has stood as the largest barrier to powerline vendors' dreams of mass-market success.

Sole entrant

Interest in home networking falls into two areas: simple home automation and PC connectivity. The former, which essentially amounts to turning lights and appliances on and off, has been largely dominated by Seattle-based X10. A number of other vendors support X10's proprietary technology, but it has no ability to carry more general data, like computer communications or entertainment.

For PC connectivity, only one company, Utah-based Inari, has actually offered a product. And the capabilities of Inari's PassPort device are, well, restrained, to put it kindly. Inari advertised the device as providing 350-kbit/sec bandwidth in raw transmission. But the firm admits that the useful rate users see is in the 125- to 175-kbit/sec range (see sidebar, "Cooking the network books").

Broadening the view to technology in the industrial market, Adaptive Networks offers a product for controlling manufacturing devices on the factory floor. The company claims actual throughput of 100 kbits/sec.

While such a rate may be more than adequate for controlling a drill press or a coffee maker, it's barely sufficient for two PCs that wish to share a dial-up-modem Internet connection. Increase the connection speed to DSL or cable-modem levels, and a user could get frustrated enough to ignore the computer and go back to the typewriters and an abacus. In the shadow of phoneline and wireless networks, which promise bandwidths up to 10- or 11-Mbits/sec, respectively, and coaxial cable, which can move data around the home at up to 100 Mbits/sec, it's not hard to see why powerline has been the poor cousin.

Retail reports underscore powerline's currently weak appeal. According to PC Data, from January through July, Ethernet accounted for 71.4 percent of networking kits sold, with phoneline taking 19.5 percent. Wireless had 4.6 percent of the sales, and powerline was dead last at 4.5 percent.

Disadvantaged

So why do powerline products suffer from relatively unsatisfactory performance, both technically and economically?

On the technical side, powerline technology must work in an exceedingly hostile environment. For a data signal, being sent over the power wiring in the average home is the electrical equivalent of being sentenced to Botany Bay, because those wires were never meant to carry data signals. "You turn on your hair dryer—[there's a power] spike," says Karuna Uppal, a senior analyst at the Yankee Group. "You turn on your microwave, and it's the same thing." Each power spike disrupts communications. "It's never a good situation for delivering data," she says.

Another factor keeping powerline technology at the back of the pack is its relative newness. Phoneline and wireless technologies have had more time to mature and establish themselves in the minds of users.

Lack of vendor participation has also contributed to the slack sales. Inari, the only company targeting consumers, is a spin-off from Novell and is too small to exert the marketing muscle that could move the mass market.

"We don't have the money to create a retail product," says Ryan Ashton, Inari's vice president of marketing and sales. Other companies developing powerline technology have stuck to either industrial customers or chosen extended development periods. Inari may have learned useful lessons in trying to sell to consumers, but the underwhelming performance of its first-generation product has molded customer expectations for the product category as a whole. "It turned into a giant for-profit beta," notes Ashton. "In retrospect we probably would have waited and launched a more robust product." However, Inari didn't do that, and now consumers associate the less-than-ideal performance with the entire category.

On the other hand, it's not as if consumers need Ethernet-like speeds for common home-networking applications. In the words of Enikia's Chylinski, 10-Mbit/sec performance is probably "overkill" for most people. "You could probably come in lower than 10 Mbits/sec and still be thrilled at the level of instantaneous response you get," he says.

Yet high-tech marketing to the public almost never adheres to logic. Computer and electronics vendors relentlessly push consumers to seek ever-higher performance, whether or not they really need it. People have become accustomed to applying such a standard, and they judge powerline products accordingly, forcing vendors to show they can compete. "That's the big issue," Uppal says. "Whoever is doing powerline home networking has to get a 10-megabit networking technology out there that is price-competitive with phoneline."

Mysteriouser and mysteriouser

Vendors claim that victory will be here in just a few months, though most remain fairly close-mouthed about the details of the actual technology that will achieve this. Enikia is using an approach "based on ISDN technology," according to Chylinski. At press time, Inari was preparing to release a second-generation product that will provide 2-Mbit/sec raw transmission, or 1-Mbit/sec of usable bandwidth, by sending data at four different frequencies. The company expects to make a third-generation 10-Mbit/sec product available sometime in 2001. Adaptive Networks will offer, in President Michael Propp's words, an "Ethernet speed" product.

Another vendor, Intellon, says it has been testing its approach in 25 to 30 homes, including new construction and a 50-year-old house, and expects a 500-home test to start around the end of the year. According to senior applications engineer David Harding, raw data rates can range from as high as almost 14 Mbits/sec to as low as 1 Mbit/sec. Harding says that 85 percent of the outlets tested notch performance of 9 Mbits/sec or better.

"You listen to the powerline vendors, and you hear a lot of stuff to the effect that they're on the verge of making a new bandwidth advance, but you never see that," says Hillary Rettig, editor of Technocopia.com, a consumer-oriented Webzine devoted to household technologies. "So they really are going to have to deliver in the near term." The lack of information on the new approaches to powerline—although understandable from the view of companies wary of their competitors—also has a familiar ring to Rettig's ears. "They basically say, 'when this comes out it's going to be wonderful, but we can't provide you with any details yet,'" she observes.

Motivation

Details are vital, because proprietary technology typically means that the only standard left is incompatibility among the offerings of different vendors. In such situations, companies typically band together through industry groups to create a common approach that will work for all. In powerline networking, though, multiple organizations are operating, each with its own biases and allegiances.

One is a working group of the Consumer Electronics Association (CEA), called R-7.3. The other group in the United States is the HomePlug Powerline Alliance, a split-off from the CEA, which includes Intel, Cisco, Texas Instruments, Compaq, Intellon, Enikia, and Panasonic, among others. Making the picture interesting is that all of the home-networking vendors belong to the CEA, but only Intellon, whose technology will be the baseline for the groups technical standard, and Enikia belong to the HomePlug Alliance.

The reason for the two companies splitting off from the CEA, Rettig says, was probably because of disagreement over how to market the powerline home-networking category. "I'm sure they were doing this because they thought they were losing mindshare [among consumers] to the phone-networking vendors," she says. Ashton takes a slightly different view. "From the beginning, it was pretty apparent that the [HomePlug] group was fashioned to promote the Intellon standard," he says.

Those two groups are hardly the end of the list. In Europe, the PLCforum (powerline communications forum) numbers members like Philips, Siemens, the German divisions of Cisco and Texas Instruments, as well as Enikia, Intellon, and Inari. In Japan, the Energy Conservation and Homecare Network Consortium, or Echonet, agreed during the summer on its own standard for home powerline networking.

But it seems that other factors must be at work. So many companies have been developing products, sometimes for years, and fussing about a technology that has yet to attract consumers worth counting. Why would corporations focus so closely on powerline networking when equivalently priced, better-performing alternatives already exist? The answer comes in two parts: the electronics industry and electric utilities.

Consumer electronics companies are keenly interested in connecting both existing devices—such as televisions, radios, and even home appliances—and new Internet appliances to a home network for online access. Phone lines aren't numerous enough to serve all those needs. And circuitry for wireless communications is expensive for an industry that perpetually faces cutthroat price pressure. The only other potential network-access point with wide enough presence to be of practical use is the power outlet. Intelligently designed equipment would require only a single cord to plug into the wall for both power and connectivity, simplifying installation for consumers. So the manufacturers would find great advantage in having a predominant standard that would let all their products work together.

"I think everything is really a potential for powerline [networking], ranging from linking two or three PCs at home to Internet appliances, low-level [appliance and equipment] control, and even eventually HDTV," says Inari's Ashton. The potential market explains why Inari is now focusing on selling silicon and intellectual property to others, including Thomson Consumer Electronics. According to Ashton, Thomson plans to release a home-networking product based on Inari technology. Because Thomson owns the RCA brand, this suggests a clear path to embedding powerline networking in prominent consumer products.

In fact, Inari says one of its main reasons for entering the consumer market was to catch the attention of larger firms. At least that's the assessment of Mark Ciesko, associate vice president of product development at Fitch, a design firm that helped redesign a second-generation offering of the Inari product. "The product was a physical manifestation, so they could point to something and say, 'This was it,'" Ciesko says. "Long term, Inari wants to be an ingredient brand, like Intel Inside." Little wonder then that one of the groups working on standards comes out of the CEA, and little wonder that the other powerline vendors primarily look to sell silicon.

The light goes on

The involvement of electric utilities at first seems more opaque. After all, they don't sell products that could benefit from network connectivity. However, whether constrained by regulation or by competition (in the United States at least), traditional power companies can grow their businesses mainly through increased consumption. The utilities are hungry for other opportunities, and they can't help but notice that they have wires connected to a vast majority of the world's homes—an asset telecom and cable companies are falling over themselves to duplicate. Enter powerline networking companies, singing a siren song that promises utilities the ability to deliver telecommunications, Internet connectivity, and entertainment over those existing power lines.

Utilities are "absolutely drooling" for technology that would offer "a fantastic opportunity for [them] to get into the telecom business," Uppal says. Some of the current powerline companies are planning trials. For example, Enikia is working with a German utility to test powerline delivery of telephony as well as E1 Internet access, the European equivalent of outfitting a home with its own T1 line. Given the often-high prices of telecommunications in Europe, the utilities could find a warm welcome.

Yet there is a long road between trials and real use. "Until you see it productized in some form and see some company purchasing it and using it beyond a trial stage, it's difficult to say whether anything will make it to market," Uppal says. "There have been a lot of technologies that have been trialed and never gone beyond there." She points to a test Nortel made of delivering telecommunications over power lines in the UK. The experiment was finally shut down because it wouldn't work with any degree of reliability. Uppal cites reliability as the "huge technical hurdle" for companies that hope to pump broadband into homes via power lines.

Prove it

In short, the future of powerline networking still has to be proven for the home, consumer electronics companies, and utilities. Cooperation among the different would-be standards groups might eliminate the possibility of competing technologies that would discourage consumers. Although efforts are being made in this regard, so far they look set to stop at non-interference, never reaching true interoperability. Technology must eventually provide bandwidth wide enough for full audio and video, not simple control signals. Most of all, powerline vendors have to overcome entrenched disbelief.

Author information

Contributing Editor Erik Sherman is a writer and photographer based in Marshfield, MA. He wrote about billing for Internet services in September.