Whither consumer robotics?

Great Debates
1

I just read this article from PC Magazine, and it helped shed some light on why the consumer/light industrial robotics industry has not taken off.

I put this in GD to ask what can be done to make the industry grow up, or if it’s possible.

This also looks like part of a greater trend that I am seeing. The US seems to no longer care about being a leader in cutting edge technologies outside of pharmaceuticals/medical/biotech research and computer engineering. I keep reading article after article about various cities striving after biotech dollars or wanting to be the next Silicon Valley (though not so much since that bubble burst). But not much of anything else. And while I grant the potential for biotech is great, I dont see enough of it to go around for all the hands trying to grab it.

And ‘high tech’ these days seems to be more about entertainment media content (video, music, games, etc.) and related delivery services than any new breakthroughs in usability of computers and other electronics.

Is venture capital still gunshy about consumer tech after the dot-com crash? Is business/finance in general so focused on the ‘service’ model now that they have forgotten about basic manufacturing?

Back to robotics. It got a push from the Roomba, and then nothing. And even that line of development seems to have stalled. I thought that commercial cleaning systems roaming office towers after work hours would be commonplace by now, but nada.

Which also leads to a ‘killer’ app an engineering friend of mine brainstormed, but that we barely see any development on either - ‘robot’ trash bins. Garbage cans that can take themselves out to the curb on trash day, or office bins that unload themselves to the dumpster at the end of the day. And its hardly an original idea.

Another app we thought up was ‘portable fridge/drink dispenser/coffeemaker.’ Instead of running to the fridge/coffeepot, have them run to us. Again, not an original idea, but commercial development? Nada. At least in the US. Japan of course, is having its fun. Follow the link to the video.

And then combine the two - a beer fridge that will recycle its own bottles! Come on, apps arent that hard to come up with - so why so little development? Where’s the Henry Ford/Bill Gates of robotics? Is the lack of standardization the biggest hurdle? If so, how do we jump it?

Where have you gone Isaac Asimov?

2

The need has to be great enough to offset the cost. Vacuuming and taking out the garbage are just not a big enough deal. Look for dangerous, time consuming or high volume applications.

3

I think the real problem is that computer scientists seem to be no closer to strong AI than they were a few years ago, and without strong AI, robots are little more than remote-controlled waldoes.

Also, power supply remains a problem. We can’t even come up with a way to reliably power a human exoskeleton. Iron Man can’t go into battle trailing a power cord, and nobody has come up with a battery or fuel cell that could power an exoskeleton for more than half an hour. See this thread, and this one.

4

I don’t think that what you are seeing is a lack of interest, it’s just that the technology isn’t there yet. Computers have been generally useful to the general public since windowing systems and the laser printer. Robotics, at best, can still only only make a foot-tall robot that can dance.

Any robot that is actually capable of any useful good, isn’t configurable. It needs to have a set input, a set output, and be in it’s own room to go through things. You can’t just set it anywhere and hope it can do something. The Roomba is the only example so far that isn’t like that, but outside of it being the first “robot” with that capability, it’s not a very useful device when you get right to it.

Biotech, on the other hand, I think is getting the short end of the stick. The whole “Flouride in Tap Water Drains your Will so the Government can Control You!”-type thinking seems to have hit the US masses when it comes to this area of research.

The US is, on the whole, rather lacking in basic science knowledge being understood by the masses. Which is something that is going to really start hurting us if it isn’t resolved.

5

I think **BrainGlutton ** has it. The problems with robotics are batteries, batteries, poor AI, power supply and batteries.

Most of our robots are RC toys. If you still need a human involved why not skip the middle man and let the man do the job?

There is also a legacy problem. Our installations are built with human operators in mind. So we need robots who are able to operate in a human environment, this brings serious inefficiencies. It would take installations that are designed for robots and not for humans.

Another problem is one of perception. We have some very seriously automated facilities (package sorting, assembly lines, factories, etc. We just don’t think of them as robots because they don’t have blinking eyes and bing-binging antennae.

Oh, and batteries.

6

There are robots everywhere. But they are in places where they can add significant value - manufacturing, hazardous materials handling, the deep ocean, on other planets. But we just don’t know how to inexpensively make a machine that can do something useful enough to justify its cost. I can see something like a Roomba making it. Give that thing a few more bells and whistles, and they might break into the mass market. Have the thing come when called to clean up a spill. We could probably manage to build that in mass market form today.

But general purpose robots are really, really hard to do. Look at the millions that have gone into humanoid robots at Honda and Sony. The idea is that if we expect robots to interact in an environment designed for humans, they’re going to have to be very humanlike. They’ve got to be able to get up into cubboards, open drawers, turn on faucets, open doors, etc. Nothing we’ve built to date comes close. Asimo and Qurio are little more than electronic marionettes, and we still consider them marvels of technology.

Oh, I just thought of a pretty cool mainstream robotic operation: Lexus LS460 Parking Guidance System. Press a button, and your car parallel parks itself. It uses a vision system to spot the open space, calculate its dimensions, and determine if it’s acceptable. If it is, the system instructs you to stop in front of the space. When you do, the car take over control and steers itself. That pretty much makes the care a robot. Add in the voice recognition system inside for controlling everything, backup cameras and various other sensors, and it’s looking more and more like a robot that you ride.

7

Keep your eyes peeled for Pleo the dinasaur robot. There’s an interesting article on it in the latest issue of Wired. Pleo will be like the robot dogs we’ve seen before but with much more realistic appearance, movement and personality.

8

Consumer robots are already making a number of inroads.

Just not in the U.S.

9

Is a CD changer a robot? an ice maker/dispencer in the fridge?

10

Perhaps the answer lies here.

11

If we’re gonna bring in YouTube, try this. :wink:

12

I dont think a strong AI is necessary for most applications. I am not expecting an R2D2 or a general purpose robot anytime soon, just smart machines able to a few tasks extremely well with mininum input once they are programmed/configured. How strong an AI was necessary for the Roomba?

Batteries and power supplies are the biggest hurdle for general purpose or completely autonomous robots, but how much power is needed to vacuum a floor or take out the trash? They do their job, then park themselves in their recharger.

I think that general maintenance will still be the springboard. It might not be challenging for most people, but it can be for retirees with arthritis and other medical issues.

And I do not know what the exact numbers are, but cleaning crews and related costs are not cheap and take up a good chunk of a property managers budget. Major office buildings are also getting ‘smarter’ and the building maintenance department now has more IT people along with their mechanics to make sure everything stays up and running. Addind a ‘smart’ cleaning system to those capabilities should not be insurmountable and could bring cost savings. Compare a crew of five techs to a cleaning crew of twenty. Get the annual cost of the robots to less than that, the business model should be workable.

The legacy issues are important also, but that crosses over into all areas of modernization. Old buildings were remodeled for plumbing, then electricity, and then again for fiber optics and IT systems - some certaintly more elegantly than others. And new buildings are designed to incorporate all those elements.

But again, I dont see any research along these lines. Standardization of hardware and the supporting software seems like the greatest hurdle as with any major technology. Property managers didnt install any of the above until they had assurances that they could get spare parts and that it would ‘play well with others’. They really dont like proprietary systems (except for their billing systems - but thats another story.)