Darpa continues to lead the technical direction of America. From the MOSIS program to this robotic challenge, they demonstrate how tapping and extending the discovery wider to everyone can drive innovation. Few grants to a few school might have included some of the programs noted in this list.
DARPA did it by hiring Willow Garage spin-off the Open Source Robotics Foundation (OSRF) to build their Robotics Challenge Simulator--like a video game, graphics wise, but optimized to accurately animate the Atlas robot performing the Virtual Robotics Challenge (VRC) tasks. Each team programmed their robot, then uploaded the program to the simulator to participate in the VRC.
BTW, the Open Source Robotics Foundation lets you download a free open source robot OS and simulator at: http://osrfoundation.org/projects.html
That is interesting. It would be more challenging when the same algorithms are loaded real time in robots and they perform the work.I guess these kinds of competitions are a great way to get the best of software and algorithms.
The biggest difference between performing a task in simulation and performing the same task with a robot is the real-world physics. For instance, some of the simulated bots navigated the mud pit by crawling through it, rather than crafting more dexterous walking algorithms. In the real world that probably won't work, since the mud will gum up the real robot's joints, gears and such.
Also, the simulation only covered three tasks: #1 Enter and drive a vehicle, #2 Walking through a building, across a mud flat and over a rubble pile, #3 Locate a fire hose, screw it onto a hydrant, then open the valve. The real-world robot contest will add five more tasks: #4 Remove debris blocking an entryway, #5 Open a door and enter a building, #6 Climb an industrial ladder and traverse an industrial catwalk, #7 Use a power tool to break through a barrier, #8-Locate and close a valve near a leaking pipe.
>> Each team programmed their robot, then uploaded the program to the simulator to participate in the VRC
Open innovation has been a triumph for ages. Yet, the innovation we see in Apple with all its secrecy in the design process shows that you may not need a great team work strategy if you have excellent minds that can work in silos. Apple makes many business cases useless because they do the opposite. And yet very successful. DARPA is hoping to see a robotics future. I do hope the big idea generator gets into it as well.
>> You said "physical robots will compete for a $2 million purse." Wow. I wonder: on what would a robot spend all that money?
Motivation. It is not the parts but the motivation to win that amount could drive the accumulation of ideas and interests that will make great leaps in the field. When the money is big, big guys take note and get ready.