The problem with robots today is that they just aren't, well ...
I have been testing out the Scooba -- the floor-mopping counterpart to Roomba, iRobot's flagship robot vacuum -- for the last couple of months.
And here's my issue with it: The Scooba involves almost as much work as simply mopping the floors yourself, but it takes longer and doesn't do as good a job.
Before starting Scooba, for instance, you have to pick up all the stray items on the floor that it might have trouble navigating around: rugs, dog beds, electrical plugs, even chairs. Once it gets going, it often struggles with hard-to-clean spots and, at least in my house, frequently runs out of juice before finishing the job.
While the Scooba is decent overall in reacting to the environment, it doesn't keep track of where it's been, which means it keeps running into the same walls and barriers over and over.
Worse yet, in a room like my kitchen -- where the floor plan isn't a perfect rectangle -- Scooba's amnesia led it to scrub some areas repeatedly while leaving other areas completely untouched.
Maybe I have high expectations, but I don't understand why such an advanced device has to be so dumb.
I grew up with R2D2, the "droid" from
. Not only did old R2 have more personality and pluck than many of his on-screen human counterparts, he could keep a spaceship flying even if it had just one wing and half an engine left to it.
To me, that's a real robot.
Now don't get me wrong, I'm not expecting an R2 unit anytime soon. I'd be overjoyed to settle on Rosie the Robot, the Jetson's mechanized housekeeper.
Instead, the best I can get is the Scooba.
Putting the Chips on the Table
I recently shared my frustrations with Scooba and discussed the current state of robotics with Helen Greiner, co-founder and chairman of iRobot.
Greiner assured me that Scooba works fine for most folks (of course) and that the path from Scooba to Rosie -- if not R2 -- is not all that far.
TheStreet.com: I was underwhelmed by Scooba. How long will it be before I can get an R2D2 or a Rosie the Robot?
I believe it will be a continuous evolution. There's a big factor you have to consider: how much people will pay for an R2D2 or Rosie the Robot.
We're progressively getting more and more features into
our domestic robots. ... We work under government development contract on more robot capabilities for navigation, manipulation, agility, autonomy and
then bring them over to the consumer side when we can invent a way to do it at a
reasonable price point.
Being able to use that success and build up features ... will get
us further than sitting down with a pencil and paper saying, "I wonder if people would buy this Rosie if I built it?" Because, chances are it will come out at a price point where they won't buy it, and it won't have the features that are really important to the users.
So, is it a cost issue or a technology issue?
about building up capability. In the '60s ... there were ... visionaries who were saying, "We should have computers like
today's laptops," and what we actually had on the market was ... just pedestrian. However ... the IBM PC and Apple, they created an industry.
Is there anything that your engineers have dreamed up in the lab and said, "Hey this is really cool, and we could do this today," but you said, "There's no market for that?"
In 2000, we had a remote presence robot running around our company. People could log in from anywhere in the world and visit iRobot ... You see what the robot sees, you can hear what it hears. ... We think it's got a huge amount of potential.
But possibly, you wouldn't pay $5,000. We actually got to the point of a prototype on that, and you know, we had to step back before we put it on the market and say, "This isn't going to fly."
On the military side, many of your robots are remote-controlled. As somebody that grew up on R2D2, my gut reaction to remote-controlled robots is that they're not really robots.
I have a little more expansive view of the word "robot." It comes from
play, and derives from the ability to do work.
I don't believe that if a robot takes commands from a human that makes it not a robot. In fact, in many scenarios, that's the kind of capability that you want to have built in. So, the Roomba is still a robot if you take a remote control and control it. It's just a user interface.
What are the next fertile fields for robots?
In elder care, helping folks remain more independent. ... Commercial cleaning: You know, we're cleaning houses today, but every store, every retail facility, every school gets cleaned every night. There's outdoors, landscaping, mowing, whether it's home or golf course or all sorts of outdoor leaf picking up, mulching, all sorts of applications there. There's construction and agriculture.
And then cars are getting more robotic themselves, first starting with safety features and then going into ... autonomy, autonomy for things running down the road and wanting to
avoid hitting other things.
Sun Microsystems (SUNW) - Get Sunworks, Inc. Report co-founder Bill Joy made a big splash a number of years ago with his worries about where technology was going, the idea that we are inventing technology that has the capacity to destroy us on its own. In terms of robotics, is that a real worry, and how far off are we from this fear?
I think in the robot case, we're so far from that. If Roomba gets out of line,
you can just pick it up and slap it.