#1
|
|||
|
|||
Robot Races
What are the current status of robot races? Saw on the TV the one in
what Nevada/Arizona, to have a unmanned ground vehicle, go some 150 miles unassisted? They got what 10 miles or so? Hum.. Small steps.. Mike I am Borg |
#2
|
|||
|
|||
Robot Races
Abrigon Gusiq wrote:
What are the current status of robot races? Saw on the TV the one in what Nevada/Arizona, to have a unmanned ground vehicle, go some 150 miles unassisted? They got what 10 miles or so? Hum.. Small steps.. Yep. BTW, it was California (crossing the border into Nevada was at about MP 135). Appearently, if you want to go slow, like the Mars Rovers, the problem is much more doable. See March 2004 Scientific American for a peak over the Red Team's shoulder. /dps |
#3
|
|||
|
|||
Robot Races
|
#5
|
|||
|
|||
Robot Races
Hum, what about using touch plates?
Or making the robot large and heavy, so that it can crush most things (sadly even the tortesies).. Or, how about combining a UAV with a UGV (Unmanned Ground Vehicles), the UAV maps out the route, using its radar and like, and the UGV follows the course laid out by the UAV? Yes, how to handle that darn bush? Touch plates or ... Mike dave schneider wrote: (mvillanu) wrote: (dave schneider) wrote: [...] about MP 135). Appearently, if you want to go slow, like the Mars Rovers, the problem is much more doable. [...] The hard part is collision avoidance and micro-navigation. How would you avoid a barbed wire fence? How would you avoid a wide/deep ditch? How do you know which bush can be run over, and which bush to go around? Yep, integrating the sensors was rather painful, it seems, including a stabilizer that introduced jitter. Note that this team chose multiple sensors, because each is good at detecting different hazards and/or different ranges, but that increased the integration problem because each presented results in a different way. Finally, to cover the course in the time allotted, you need to average a minimum of 19.8fps. So all of the micro-navigation capabilities of the robot must be able to work extremely quickly, AND get the robot to react in time. In a way, getting UAV to land by itself is a lot easier. I would think you'd need a combination of PIDs and expert systems, gluing them together with fuzzy logic. /dps |
#6
|
|||
|
|||
Robot Races
How would
you avoid a barbed wire fence? How would you avoid a wide/deep ditch? How do you know which bush can be run over, and which bush to go around? Consider the cockroach navigating at night. Insect antennae. Long and light articulated fiberglass poles. This slows one down but keeps one safe like a blind man's cane protects him. ^ //^\\ ~~~ near space elevator ~~~~ ~~~members.aol.com/beanstalkr/~~~ |
#7
|
|||
|
|||
Robot Races
Abrigon Gusiq wrote:
Hum, what about using touch plates? Or making the robot large and heavy, so that it can crush most things (sadly even the tortesies).. Excluded by the rules of the race. Or, how about combining a UAV with a UGV (Unmanned Ground Vehicles), the UAV maps out the route, using its radar and like, and the UGV follows the course laid out by the UAV? Yes, how to handle that darn bush? Touch plates or ... At the speeds required for the race, touch plates would be way too late in warning the vehicle. Even the fibreglass "antennae" suggestion would be difficult to do for much more than a crawler. Note that the UAV suggestion is somewhat like what the Red Team did: they used GPS and terrain maps to pick out useable courses ahead of time; when the race stewards released the CD of the route information, they did an all-nighter to tune their closest course, and then the sensors on the vehicle picked up the stuff too small to show on maps. /dps |
#8
|
|||
|
|||
Robot Races
|
#9
|
|||
|
|||
Robot Races
Abrigon Gusiq wrote in message ...
Hum, what about using touch plates? Or making the robot large and heavy, so that it can crush most things (sadly even the tortesies).. Or, how about combining a UAV with a UGV (Unmanned Ground Vehicles), the UAV maps out the route, using its radar and like, and the UGV follows the course laid out by the UAV? Yes, how to handle that darn bush? Touch plates or ... Mike I thought about the UAV also. But then it must be able to fly out to a large enough radius in order plot a long enough route for the vehicle. I don't think vehicle can cover the course in time. The UAV can probably solve the wide/deep ditch problem, but there's that barbwired fence and bushes to deal with. Touch plates are a good idea for the bush, but from personal experience, the speed at which the vehicle must move in order to finish quickly winds up upsetting the bush. |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
How to Mars ? people / robot debate | Dan DeConinck | Space Shuttle | 7 | January 24th 04 08:16 PM |
How to Mars ? people / robot debate | John Doe | Space Station | 1 | January 16th 04 09:21 AM |
Humanities, engineering contribute expertise in developing NASA'sRobonaut (Forwarded) | Andrew Yee | Space Station | 4 | December 5th 03 07:39 PM |