Forty-three years after Rosie the Robot first made dinner for the Jetsons, the field of consumer robotics is still struggling to find its groove. Robots have become commonplace in the manufacturing industry–and the military’s deep-pocketed interest is growing–but only a handful of commercial robots have made it into the homes of ordinary people.

What’s going on?

“You need four things to make a good robot,” said Thomas Parrish of Openware Robotics. “It has to be useful, reliable, simple to use and affordable.” But the research that that requires is costly, and the price points of commercial robots thus often wind up higher than most consumers are willing to pay. At $2,000, Sony’s Aibo robotic dog is a mighty expensive source of amusement.

At the 2005 RoboBusiness conference in Cambridge, Mass., at a panel on personal robots, the possibility was also raised that roboticists were out of touch with the needs of the general consumer. In an industry still dominated by geeky researchers, Parrish grumbled, “a lot of roboticists are trying to reinvent the wheel.” Could larger companies like Johnson & Johnson (nyse: JNJ – news – people ), Kimberly-Clark (nyse: KMB – news – people ) or Procter & Gamble (nyse: PG – news – people ) provide them with direction?

A P&G representative sitting in the panel’s audience raised his hand and noted that his company could identify at least 20 problems that consumers would like to see addressed around the home. Several panelists nodded their heads. “We also know about how much money most people are willing to pay for the solution,” he continued. “It’s around $20.”

One approach that several companies are trying is the manufacture of so-called PC bots, which are made from parts of personal computers. “Because the parts are readily available,” explained Thomas Burik, chairman of White Box Robotics, “it shouldn’t cost much more than a high-quality laptop.”

And what does the PC bot–which looks suspiciously like a bigger, bulkier hard drive–do? “Oh, anything,” he said. “We imagine that these will be like PDAs, where the user can write programs to customize them to do whatever they want.”

Other conference attendees weren’t so convinced. “Servers on wheels,” scoffed one. “They don’t even have arms.”

And arms–and legs–are a key part of the runaway success of Robosapien, a 14-inch humanoid bot made by Hong Kong-based toymaker WowWee.* Robosapien, which does little more than walk, dance and speak “fluent international ‘caveman’ ” (it also belches), took $1 million to develop. But since its introduction in early 2004, Robosapien has sold 1.5 million units at a retail price of around $100. The company expects to sell 3 million units by the end of the year.

Indeed, throughout the conference, low-tech, low-cost consumer robots were hailed as lessons in success. IRobot’s Roomba robot vacuum cleaners, which sell for $300 or less, have sold more than 1 million units in the past two and a half years. Roboticists gripe about Roomba’s less than sophisticated technology, but as Joe Engelman, the retired roboticist who is generally credited as the father of the field, puts it, “You can’t argue about their success.”

Dan Kara, president of Robotics Trends and the chairman of the RoboBusiness conference, points out that 25 to 30 venture capitalists attended this year (last year’s event drew only four). He estimates that the market for personal and service robots will reach $4.2 billion next year.

Meanwhile, Roomba’s co-founder and chairman, Helen Greiner, is also bullish about the future of her market. Robots in the home? “It’s not a question of if, it’s a question of how many,” she says firmly.

More here.