DARPA Seeks More Efficient Robots

[GeekTinker]

http://www.informationweek.com/news/government/info-management/240003188

Agency challenges innovators to develop technology that improves robotic mobility and control. 

By Patience Wait   InformationWeek
July 05, 2012 11:00 AM 

Robots are about two orders of magnitude less efficient than humans and animals in using energy to produce motion, which limits their usefulness for military and other applications, according to the Defense Advanced Research Projects Agency (DARPA).

To address this limitation, DARPA has issued a solicitation to industry for help in developing "actuation" technologies--those involved in a robot's movement--with a goal of achieving a 2,000% increase in power transmission efficiency.

DARPA's Maximum Mobility and Manipulation (M3) program, inaugurated in March 2011, is aimed at developing design tools to improve robotic performance, improve fabrication processes and control methods for mobility and manipulation, and produce prototypes. 

The M3 actuation project expands on DARPA's robotics R&D efforts. "By exploring multiple aspects of robot design, capabilities, control, and production, we hope to converge on an adaptable core of robot technologies that can be applied across mission areas," Gill Pratt, DARPA's M3 program manager, said in a written statement. "Success in the M3 Actuation effort would benefit not just robotics programs, but all engineered, actuated systems, including advanced prosthetic limbs."

The solicitation outlines two goals. In one, bidders will be asked to develop and demonstrate high-efficiency actuation technology that lets robots similar to those using the DARPA Robotics Challenge's government-furnished equipment platform have 20 times longer endurance than is currently possible (only 10 to 20 minutes), or up to 400 minutes. Companies will be expected to share their design approach at DARPA's Robotics Challenge in December 2013 and demonstrate their systems a year later. Awards will be limited to $2.5 million per project per phase.

In a second track, bidders will be asked to conduct basic scientific and engineering research into improving the efficiency of actuators, without requiring that it be applied. Those awards are capped at $500,000 per project per phase, or $1 million maximum.

The Office of Management and Budget demands that federal agencies tap into a more efficient IT delivery model. The new Shared Services Mandate issue of InformationWeek Government explains how they're doing it. Also in this issue: Uncle Sam should develop an IT savings dashboard that shows the returns on its multibillion-dollar IT investment. (Free registration required.)

[Scott A. Kampas]

KurzweilAI.net had a nice piece on this, too:
http://www.kurzweilai.net/darpa-seeks-2000-percent-increase-in-robot-power-transmission-efficiency
.

And some other cool things recently:
http://www.kurzweilai.net/could-plasma-light-extend-moores-law

http://www.theregister.co.uk/2012/07/04/quantum_computing_room_temperature/

Scientists Twist Light to Send Data: Beams of Light Can Be Twisted and
Combined to Transmit Data Dramatically Faster
http://www.sciencedaily.com/releases/2012/06/120625133349.htm

[gswr...@gmail.com]

Scientists Twist Light to Send Data: Beams of Light Can Be Twisted and
Combined to Transmit Data Dramatically Faster
http://www.sciencedaily.com/releases/2012/06/120625133349.htm

Just WOW! Where I work, we send 80 OC-192s down a single fiber, and in the commercial world, I think the theoretical maximum is ~180 OC-192s down a single fiber. That's about 1.8 Tbps. These scientists may enable us to send TWICE that in either a single channel, or 8, depending on how the wavelengths affect each other. They don't explain in the article if all 8 beams are of the same wavelength.

That could be a potential 400+ Tbps . . . Down a single fiber. Just Wow!