I just picked up a copy of The Personal Robot Guide (published 1984) a few months back and was amazed by the wide variety of robots once in development. These little guys would “talk” via voice samples, serve you drinks, roam around the house, turn lights on and off; just about anything cool.
These days, however, we’re seeing pretty much no progress in that area. What happened to the world’s interest in robots? To tell it from the available data in the mid-80s, we’d all have our own Johnny 5 (of Short Circuit fame) equivalent by now.
I realize that Japan is still working on robot-y projects (though they don’t seem to do much of interest) and there’s that “Roobma” thing on the market that vacuums your floors, but those are neither here nor there. Why have robots fallen so out of favor?
Erm… Ever seen a car assembly line?
How about bomb squad robots?
Or Mars probes?
Now, if you’re talking about androids rather than just “robots” than you have to consider that they pose great technical challenges for very little practical value. In other words, it’s not really worth investing in them other than to show off your technical know-how, like Honda and Sony are doing.
Actually, robotics is a pretty big hobby these days. Aside from research robots, there is a large hobby robot industry. People building robots for “Battlebots”, personal robots like the Sony AIDBO, The myriad robot toys for kids, and the lego Mindstorms.
Moving up scale a bit, there is going to soon be a rather large market for home robots, I believe. Have a look at this robot from Sony:
This robot is amazing. You have to see the videos of it in action on that web site. Word is, these things will be for sale in maybe 2006 or 2007. Under 20 grand, at first. Clearly, the first people who buy it will be wealthy early adopters and maybe robotics research departments and such.
But every couple of years there will be another generation that will be half the cost and twice as good. A decade from now these things may be as common as computers.
I don’t believe making intelligent decision is part of the definition of what constitutes a robot. Bomb squad robots and Mars rovers are mechanical devices that perform relatively complicated tasks with some level of cutomisation. Automatons can only perform a single, repetitive task.
It all depends on how you define “robot”.
I have a hard time seeing that happen. Robots like Qrio don’t do anything useful. You can do a lot with computers, which is why they’ve been commercially succesful. Qrio and before Aibo are cool. For about 20 minutes. Then, you realise they don’t really do anything. They lack a selling point.
What I do see as potentially succesful are robots like Electrolux’ Trilobite. Things that are designed with some sort of purpose in mind, be it practical or entertainment-related.
Oh, I think Qrio could be plenty useful. His wireless LAN connection and web browsing ability might allow him to answer questions like, “How hot is it outside?” or “What’s playing at the Cineplex 12 tonight?”.
Being mobile means he may be able to fetch small items.
He’ll be supremely entertaining. He’ll do a dance for you. He’ll download songs from the internet and play them while dancing. He’ll be a walking telephone. Imagine sitting on the couch and going, “Hey Qrio - call Bob for me.” And a few seconds later Bob answers through Qrio’s speaker.
Qrio will be able to play games. Eventually, you’ll be able to play chess with him at a real chess board, and he’ll pick up and move the pieces just like a human.
He’ll be a great sentry. With a video camera and a wireless video connection, you can send Qrio to check out the house, and watch on your PC as he goes from room to room.
And I’m sure there will be tons of applications we haven’t thought of yet. Just like the Aibo is being used for a lot of things Sony didn’t imagine.
These things have to drop to under $3,000, and gain a couple more generations of reliabillity and functionality, and they’ll fly off the shelves. After all, Sony sold thousands of Aibos for $2,500.00, and they don’t have 1/20 the cool factor or capability as Qrio.
The problem I see is that my cell phone can already most of that and it fits in my pocket. Plus, I got it free when I subscribed.
I’ve seen Qrio and I’ve played with Aibo and I can assure you that their dancing gets pretty boring real soon. Certainly not worth 3000$. Also, the Aibo wasn’t exactly a resounding success - not for a company like Sony.
How do you sell a product like Qrio? Computers made their way into our homes in great part because people were using them at work. I don’t see that happening with androids. Ultimately, for something in that price range, practicality will always win over coolness.
You can find quite a lot going on among hobbyists, especially now with mini-ATC PC’s available that will fit into a 7"x7", 5" inch high box, and micro PC’s that are even smaller. I understand that there is a considerable argument over if the PC should be built-in to the robot and the whole unit fully self-powered and mobile as convention expects, or if the PC “brains” should be stationary instead: have a conventional stationary PC somewhere running the “robot” program, and a mobile vehicle with sensors, and an RF link between them. So the “robot” program is running on the PC, but it is getting info from and guiding the remote vehicle just the same.
…
But really, what do you do with it? The main problem is finding a use that most people would agree is a good idea: vacuuming a floor kinda makes some sense, but not a lot else really does. One task they used to say was really nifty was the ability for the robot to wander your house as a sentry when you weren’t there, watching for intruders–but honestly, simply intalling a wired conventional alarm system with IR sensors is really a much cheaper, easier and more dependable way to acomplish the same thing (and with wireless PC tech heading where it is, the “wired” aspect of these alarms will soon disappear–making “installing” them even easier).
So really, one needs to find a task that robots can do in the home–and do better enough than any other solution–to justify their costs.
~
Many robots (pictures & descriptions) can be found at the Robot Menu. Most are autonomous (i.e., not remote-controlled toys), and a disappointingly small number have been designed to go on murderous rampages.
