Rare Desert Walking Robot: Mojave or Bust

An anonymous reader writes "Robust walking robots are still surprisingly rare. The Astrobiology Magazine is reporting today on the German-American Scorpion Project to conquer 25 miles of targeted navigation into the Mojave Desert and back autonomously. The eight-legged robot is triple-jointed and must travel by day (solar-batteries) for two-weeks alone without human intervention. Because it's a scorpion, the camera is in the tail."

Comments

Re:What the heck did they build that out of?

[moonbender]

I'm not sure, but perhabs the article wrote cm when they meant mm. That'd make more sense.

Re:8 legs really needed?

[AndroidCat]

It is a rugged environment. Perhaps they planned in case one or more legs gets gunked up? (Good software if it can cope losing random legs.)

On Mars mission, it's hell trying to get a repairman in, especially on weekends! (And the rates they charge!)

Re:8 leggs really needed?

[Anonymous Coward]

I have a couple of spiders in my backyard that would like to demonstrate to you how handy all eight legs are in their daily goings-about.

Bust, probably...

[cr0sh]

My first impression after seeing the robot on the site was "what the hell are those wires doing out in the open?". Given the conditions of the area they are doing the run in (where daytime/nighttime temp fluctuations are large), exposed wires would be in for a real beating. Even if well protected, with metal sheilding or whatnot, they still will be flexed a lot - here's hoping they have an efficient AI that can change tactics when a leg motor fails.

I also question when they will be doing this run of 25 miles. If they don't pick the right time of season, they will run into either dust storms (strong enough to rip the paint off your car and needing a new windshield), thunderstorms (dropping a lot of water in mere hours), or a combo of both if they are really unlucky.

This robot is going to take a few days to go the distance. It will have to deal with desert conditions (both day and night conditions, and weather) and desert obstacles (brush, trees, loose soil, rocks, animals, washes, etc).

I really question the whole project - was the goal to make a legged vehicle or to go a distance autonomously?

I think about the DARPA Grand Challenge, and I think what would be the best vehicle for such a thing. Then I thought it would be a challenge to go any distance autonomously. For the terrain, legs are OK, but use a lot of power. Considering the obstacles likely to be encountered, a much better and more efficient system would seem to use wheels.

I would go for a system similar to what the Mars Rover used, the suspension system with six wheels, but scale it up a bit to use larger ATV wheels. Keep the engine, but throttle it back greatly (under computer control) to maximize fuel consumption (properly tuned, etc with proper throttling, you can get hundreds of miles per gallon of fuel). Keep solar panels and batteries on-board, which at minimum could run the system in the event of an emergency to "call home". Such a system would be able to make a 25 mile run in probably a day or two maximum, and could possibly compete in the DARPA challenge.

At the end of the day, though - regardless of whether this legged wonder does it, or somebody builds another device, wheeled or legged, that does it, even over a much shorter distance, they will have come up with a solution, and proved that solutions exist, to several "grand problems" in autonomous robotics that would prove useful in a number of other situations (ground rescue, factory work, planetary rovers, underwater exploration, etc).