It looks like a cross between a computer game and the old slot car race sets.
See <http://youtu.be/o38DpmXnZn8> and <http://youtu.be/M-iCj7xRm78>. I haven't
got my mind fully around it yet, but it seems like it would be an interesting project to
build something like this with the EV3. (It would be a quantum leap over my NXT
racers.) I think the EV3 race car robots would be slower than the Anki cars, and thus maybe
not as thrilling to watch, but certainly fun to build. Of course this is a long range project
as I am months away from an EV3 and longer still, until I understand the EV3.
But, to start planning, a question about the EV3 IR beacon - the spec says it has 4 channels,
does this mean that if 4 robots each had a beacon set to a different channel, could a robot
with an IR sensor uniquely identify the other 3 robots? Thanks for any info about the EV3 IR,
and any thoughts on the Anki Drive.
I'd venture to guess that an EV3 robotic car would be slower just because of the weight of the EV3 components. You can't accelerate as quickly and nimbly when hauling a big load. Nonetheless, if you can figure out the algorithms, I'm sure that the EV3 brick should be able to respond in a similar way to the anki cars.
hi,It would be a quantum leap over my NXTracers.
exciting, a very interesting idea indeed.
But first a remark: a quantum leap is the smallest change at all that is physically possible in theory and practically.^^
I think it would be lot of more than just a quantum leap ;)
And second, I'm quite optimistic that the ARM9 provided by genuine ARM machine code will have enough power to build such kind of racing cars.
Actually it might be not exactly "racing" what they'll do, but maybe creeping or anything in between. :D
When running partially autonomously, additionally some real fine super-smart-inertial IMUs featuring on-board motion and localization processing would be needed (like e.g. the MPU-9150), and I think you'll also need 2 or 3 distinguishable external reference IR beacons and 4 Sharp GP2D12(0) IR distance sensors or, alternatively, 4 ultrasonic sensors (in single-shot-mode to avoid multiple echo noise), all mounted to have overlapping detection angles like, e.g., facing 90Â° left, 30Â° forward-left, 30Â° forward-right, and 90Â° right.
Pathfinding is not much more than a bug2 algorithm (real time processing) to be implemented - the rest will be not just peanuts but absolutely a solvable problem IMO.
interest in doing it. Regarding where the intelligence is in the Anki racers, I wasn't sure - in the race car?,
in the iPhone?, or in the human holding the iPhone? And thinking a little more about such a project, I'd
like all the intelligence in the EV3 on-board the robot, and the external data to come from sensors on the
car. Then there is the question of whether the cars communicate with each other, or whether they just
react to the location of other cars nearby. But I think my first step is to wait for an EV3 and see how fast
I can collect what data with the new sensors. I'd like to stay with all LEGO parts, sorta' a self imposed artificial
if the iPhone was on board the cars, they might use the cam for pathfinding and localization - if not (what actually appears to be), there must by sort of comm unit for sending sensor values and receiving drive (acceleration, brake) and steering commands from an external iPhone. I'm thinking of the iPhone as a kind of Wii (You) reMote for SuperMario Cart plus remote - real-time localization in the course track and of other vehicles by what-ever kind of sensors mounted in the car (IR? Transponders?).
Users browsing this forum: No registered users and 6 guests