META: A robot with Triangular Tracked Wheel locomotion mechanism

[mheljo88]

Well , Nice Robot but Its Kinda heavy , :) Well i guess thats where the Triangular Wheel is used for

[aoshima6566]

This rover would have no problem at all driving through a field of large jagged rocks. Good work!!

[nonbarbari]

loool funny to see 13 ppl that are jealous of robot haha... Tum up dudes, keep it up

[RockzoPhD]

Boosh!

[tekinak007]

You dont design high end robotics with wireless to start with idiot. Adds complication and effort to the project when its not even required to be at that stage in production/testing.

[tekinak007]

Not to mention power needs, this way you can test without having to worry about flat batteries.

[j00d00d]

Talk about reinventing the wheel....

[amirhst]

Thanks, it is so ;-)

But honestly, I don't agree about your comparison of robustness between tri-star and triangular systems!

I should remind you that this is a tracked system while tir-stars are wheeled ones.

We've tested this robot on a 50 deg ramp so many times. It moves up, down and even rotate pretty good on that ramp where I'm sure no tri-star system has enough traction (due to the contact surface) to go up!

[oateslarsen]

I looks cool, but I don't see the practicality of having the tread bases also act as wheels...

[amirhst]

well, it seems there is no argue about its tracks so ...

when the robot rotates its triangular frames so that they stands on their corners it's like a differential steering four wheeled robot and when it rotates its frames continuously it moves some how like a four legged robot.

I suggest having a look at my other video entitled as "AriAnA Rescue Robot Team"

[predben]

getting over large rocks probably

[xskywarpzero]

some robots have similar tread spins that climbs up and down stairs

[DocWolph]

Impressive. This could be topped only with a legged transport. This could be a Moon or Mars rover.

[VulpisFoxfire]

Hmmm. Out of curiousity, any reason you use the noisy and kinda thumpy tread system here, instead of the similarly designed Tri-Star wheel arrangement?

[amirhst]

to benefit inherent capabilities of Tracked vehicles (eg. less ground pressure, grater traction force by a given torque ...)

[siyamand]

eyval be bachehaye khodemoon dametoon garm mokhe bache haye ma japoniyaro mizare jibe kocholoye agabesh......proud to be IRANIAN...:)

[makenroug]

valienteeee mierda tu robott

[lukasvip7]

Ugly maschine

[amirhst]

Since this project is supported by industry, we need their permission to publish any material and I'm so sorry to say that we've presented this work solely in some Persian conferences till now.

[amirhst]

Certainly Not!

The robot corrects its orientation over the time using mentioned algorithms (of course not in this video :D). It may have at most 5 deg. error in very complicated situation. Relying only on dead rockoning won't have such an accurate result!

[amirhst]

Not at all. The IMU fuses output of these built in sensors to deliver more reliable orientation data at a dynamic condition.

[amirhst]

3) As far as I know, most IMU's have 3 gyros, 3 accelerometers and one digital compass to measure 3D orientation and 3D accelerations (3 linear accelerations as well as 3 rotational accelerations)

[amirhst]

2) We use state of the art Industrial DC motor controllers (Mars rovers also use the same motor+controller combination). These motor controllers utilize encoder outputs for velocity and position controlling.

[amirhst]

If it helps, we actually use Maxon (maxonmotor[DOT]com) products in the powertrain.

[amirhst]

1) Yes, we do it. To decide about where and how robot should go, we use exploration and navigation algorithms which are based on an Occupancy Grid map. This map itself is generated automatically by a Simultaneous Localization And Mapping (SLAM) algorithm using scanning Laser Range Finder (LRF) , Inertial Measurement Unit (IMU) and shaft encoder data. In out door applications, we add stereo vision and GPS data to aforementioned collection.

[amirhst]

Thank you very much for your kind notation.

In this video, the robot actually was controlled using an ON/OFF switch pad (not position or velocity controlled).

Now velocity of tracks and position of Triangular frames are controlled preciously by accurate motor controllers. Also robot orientation is monitored using an IMU.

[idrobinhood]

might as well send it to Mars, moon, titan as a rover to do research. Might as well start programing some movement and direction patterns so to remote and able it, an to move at further distances as well with out wires attached to it..

[amirhst]

It's exactly what our team have worked on during these 2 years :-)

I hope I can be allowed to put some newer videos of current state of the project.

[0ptimus89]

ok its cool but is it a real robot i mean u are operating it with a remote control shouldnt it be without one

[amirhst]

Thanks for your interest on our work.

As title implies, it was first examination. Now the robot uses WiFi to communicate with control station but, it still needs operator's aid for high level decision making (e.g. exploration).

As far as I know, all "rescue robots" are tele-operated systems and no fully autonomous one has been demonstrated in real situation yet.

It is hard to believe that the published reports after 9/11 insist on "tethered communication" between robot and control station.

[vuquyhungbk]

I used five motor. four ones for four tracks. another one to crawl the rear one and rescue robot from overturn situation. My robot can stand by himselfl if the body is flipped upsidedown

[amirhst]

Self righting is a grate capability in mobile robots, but it rarely is useful in practice. ‎For example iRobot packbot platform very efficiently turns itself when turned over, ‎but the story about packbot with a manipulator (or navigation sensors) is completely ‎different.

That is why we try to design a robot which resists against overturning.‎

[vuquyhungbk]

ok, I know what you did. that good idea. lastime, I confused

[amirhst]

You're welcome!

The misunderstanding seems was due to your concentration on your own design ‎which I guess uses only one motor to drive tracks and frames. If I'm true, how many ‎motors have been employed in your robot?‎

[vuquyhungbk]

it is quite difficult to understand. you said:"both front triangular frames are controlled by one DC motor and both rear frames by another one; "I understand that: two front tracks are controlled by one DC motor and two rear tracks are controlled by one DC motor.

1. how can you steer your robot that it need the different velocities.

2. one motor for front ones, one for rear ones. totally, your system needs only two actuators.

I really confuse. Pls, make it clear for me.

thank in advance

[vuquyhungbk]

hi, I interseted in your robot. i look quite powerful and flexible. I wonder that how many actuator was used.

[amirhst]

Thank you!

This robot uses TTW (Tracked Triangular Wheel) locomotion mechanism. In this mechanism, both front triangular frames are controlled by one DC motor and both rear frames by another one; also each side tracks are driven by a single DC motor. So we need 4 motors to control entire system.

[sigmamgis]

Interesting robot!

Would you please let me know how can I get more information about this robot?

[amirhst]

Thanks!

This project is done at Robotics Research Laboratory of IAUCTB and I'm one of the team members.

We haven't published more information about it yet and I hope to do it soon.

Anyway, META is a modified version of our older robot, ARIAN (take a look at my palylist) which itself was inspired from Space Cat. (see again my playlist).

[amirhst]

I think this isn't the exact place your and my taxes go nor we can be proud!