May 19, 2003
Useful Home Robots For Lawn Mowing And Vacuuming

When my brother recently told me how much he likes the $200 Roomba home robot vacuum cleaner it served as a reminder that we are living in the era when, after decades of science fiction speculation, personal robot servants are finally starting to make an impact on everyday life. In another sign that robots are beginning to invade the home front the Toro iMow robotic mower is available for about $500.00.

Friendly Robotics of Israel makes a similar product for about $600 called the Robomower.

Robomower stays within the boundaries of a property using a thin wire perimeter placed on the edges of a lawn -- something akin to the invisible fence sensor technology used to contain pets.

The Robomower's brain records the lawns dimensions and obstacles, then operates on a computer-run navigation system until the job is done.

There do not appear to be (unless you dear reader can point to any others) any other major useful types of home robot products at this point in time.

While less useful from the standpoint of saving labor, Sony continues to come out with improvements for very popular the Aibo robot dog.

Also, in conjunction with AIBO's fourth anniversary, Sony will be releasing a sleek, new cobalt-blue colored AIBO Entertainment Robot [ERS-210A/LI] dubbed AIBO Cyber-Blue. Both AIBO EYES (suggested retail price $119.99) and AIBO Cyber-Blue (suggested retail price $1,299) can be purchased by visiting the Sony Style Stores (New York and San Francisco), Sony Gallery (Chicago), select retail outlets nationwide or by visiting http://www.us.aibo.com beginning May 20.

With AIBO EYES software, the four-legged robot can now be controlled remotely via e-mail commands. Users can send an e-mail message to the robot and receive a JPEG image with their computer or other mobile communication device, capturing a picture of what AIBO sees allowing them, for example, to view their home or children while away.

In addition, AIBO EYES software will allow family and friends to communicate through audio messages. For instance, AIBO owners can now send a message command from a remote device, such as a PC or mobile communication device, to AIBO and have the robot deliver a pre-recorded message aloud such as congratulations! Further, AIBO EYES will also enable owners to remotely e-mail message commands to AIBO and have the robot perform selected songs, such as When the Saints Come Marching In and Ode to Joy.

Speaking as someone who loves real biological dogs the idea of getting a robot dog strikes me as somehow disgusting. But Aibo has sold over 100,000 Aibos. Go figure.

One use of the Aibo robots is to compete in robot dog soccer leagues run by academic groups to test new software.

Rigged for soccer, the dogs communicate by radio and are fitted with tongue-depressor-shape "memory sticks" that slip inside the dogs' abdomens and contain all the team's codes. In Carnegie Mellon's software, dogs trade off being the leader depending on which one is closest to the ball. The leader goes for the ball, while teammates deploy themselves in support positions.

Robotics researchers find that competing to develop better soccer robots is accelerating the development of better robotic software.

Robotic soccer has been embraced by many researchers in the fields of artificial intelligence and robotics because it forces them to find ways for robots to learn how to work together and to do so at real-life speed.

There is even an International RoboCup competition that will be held in Padua Italy this summer.

Carnegie Mellon researcher Hans Moravec sees home servant robots with larger repertoires coming around 2020.

By 2020, "universal robots" will arrive. They can prepare an egg, put out the dishes and clean the table. At this stage, robots execute applications with "reptilian inflexibility," Moravec says, unable to deal with the unexpected.

My guess is that some of the tasks envisioned to be performed by home robots will no longer even need to be done by the time robots that can do them become available. Take the ironing of clothes for instance. It seems far more likely that nanotechnological developments of new textiles will result in clothes that do not need ironing and which can be cleaned far more easily. Imagine nanotech clothes that can be switched into a mode where they become repellant to dirt. You might pass clothes thru an electromagnetic field and as the clothes dropped thru the field the dirt would all come off. Other nanotechnological advances could lead to the development of self-cleaning shelves and carpets that transported the dirt to dirt collectors.

Moravec does not expect to see artifically intelligent robots that can engage in human-like abstract reasoning until 2050. Therefore, we have decades to live yet before we have to start worrying about robots taking over the world.

Joel Burdick, Professor of Mechanical Engineering at the California Institute of Technology and member of the Cal Tech Robotics Group, sees five technological obstacles to the greater use of robots.

According to Burdick, robots will become a part of every day life once five technological hurdles are cleared: computing power, sensor technology, power supply, motors, and smarts.

The first 4 obstacles that Burdick cites will all gradually fall because of technological trends that are being driven by much larger existing markets for other types of products such as laptop computers, cars, and consumer electronics devices. Microprocessor speeds will keep advancing because Moore's Law has at least several years to run yet. Even as current lithographic techniques for making semiconductor circuits reach their physical limits nanotechnological approaches to making electronic circuits will probably allow processors to keep getting faster. Newer and better generations of sensors will also continue to be made. A lot of work is going into better power supplies to satisfy the demands for a large assortment of consumer electronics and transportation applications. Motors similarly have many other existing uses.

In the long run software is the main class of problems that the robotics researchers will need to solve. Most of the other problems will be solved by scientists and engineers working to make better products for a wide range of applications. The biggest question in the long term is whether the development of better algorithms or the development of faster processors will be the rate-limiting factor for the development of artificial intelligence.

For more reading on robots the Robotics.com site has a nice page of links to sources of information about all aspects of robotics.

Share |      Randall Parker, 2003 May 19 08:09 PM  Robots Home


Comments
Post a comment
Comments:
Name (not anon or anonymous):
Email Address:
URL:
Remember info?

                       
Go Read More Posts On FuturePundit
Site Traffic Info
The contents of this site are copyright