The DARPA Robotics Challenge has started and the requirements push the robotic envelope (Fig. 1). The original DARPA Grand Challenge was almost a decade ago and the first pass was not too encouraging. The next iteration was more successful. These days driverless robot cars are moving onto the regular roadways (see Are Safe Robot Swarms Possible?).
DARPA is looking to award $2 million dollars. That may sound like a lot of cash but it probably will not even help the challengers to a breakeven point. The cost of a Willow Garage PR2 robot (see Cooperation Leads To Smarter Robots) that could do some but probably not most of the challenge is more than a quater of a million dollars.
I also noticed the Men's Health Electric Car Challenge. This is a bit less challenging but probably as entertaining. The challenge is actually to drive a 2012 Ford Focus Electric cross country (Fig. 2).
The Men's Health Electric Car Challenge is going on right now. It highlights the current state of technology and will do a bit for raising awareness of electric vehicles.
The DARPA Robotics Challenge is just releasing the goals and specifications for the challenge. Like the DARPA Grand Challenge, it sets goals that may not be attainable but even a partial success by any of the participants will be a big step forward.
The DARPA Robotics Challenge is designed to address "ground robotic capabilities to execute complex tasks in dangerous, degraded, human-engineered environments." It will consist of three key events: a Virtual Disaster Challenge, and two Disaster Response Challenges. Some of the tentative goals include:
- Drive a utility vehicle at the site.
- Travel dismounted across rubble.
- Remove debris blocking an entryway.
- Open a door and enter a building.
- Climb an industrial ladder and traverse an industrial walkway.
- Use a tool to break through a concrete panel.
- Locate and close a valve near a leaking pipe.
- Replace a component such as a cooling pump.
The actual details are subject to change. Like the Grand Challenge, the particpants need to address a range of issues from safety to performance.
The challenge is not simply a matter of getting from point A to point B and avoid a few obstacles. The robots will be autonomous but interact with people in more advange mode than just assuming they are objects to be avoided. There is likely to be a requirement that non-expert operators will have to be able to instruct the robot(s) to perform tasks. This would be on par with an emergency situation where a human operator needs to prioritize or specify tasks to be performed.
Proposals are due by the end of May. Some may get funding up to $3 million while most will likely use some of their own funds. The challenge will address hardware and software and it is possible for some teams to use provided hardware and deliver software to address the challenges. The project is extremely difficult and it will be a pleasant surprise if any of the participants completes a significant portion of the challenge.
I'll be watching both challenges.