Go Down

Topic: Need some professional advice on my project ~~ (Read 1 time) previous topic - next topic


i am so dead for my final year project.

Well your supervisor wants a good kicking for letting you attempt this.

any idea on how i can simplified it and get a more or less the same system that i wanted

Not really it is a complex system, getting a knife wheedling robot arm is a bit on the tricky side.
How about a robot arm that speared cheese on the end of a cock-tale stick.   


well, actually I would not consider it all that impossible as others are saying. It just depends on your setup. you need to change the environment to make your task easy

a) tracking

say you have a white surface, and three different fruits with distinct colors. You mount a webcam on the ceiling, looking straight down. You track the centers of the colors.

b) gripping & cutting
think of it as an x/y plotter: (something like this: http://makerbot-blog.s3.amazonaws.com/wp-content/uploads/2012/03/nyc1591-650x437.jpg)

the "arm" consists of a gripper with an integrated knife, mounted on a linear bearing system for movement in the x axis, and a linear actuator for movement in the y axis.


you put 3 fruits on the table, which all have similar size (fit for the gripper) and different colors. You want to cut the red apple? The software on the computer finds the centre of the red blob from the webcam feed. The "arm" moves its x position to the blobs x position. The "arm" expands until the linear actuator has positions the gripping element on the blobs y position. Then the gripping element carefully closes. The cutting element cuts it in two pieces. Then the whole thing retracts and voila.


And yes. you can do that with roborealm and arduino.

And no. even what I suggest will not be easy.


thank fkeel for the input ~~

is making anything like this difficult ??


no its not.

I actually did something very similar to that for my first arduino project. (I take it its a webcam mounted on two servo motors.) ... Sadly that was before I started documenting my work or I would send you a link.

If I may suggest. You will learn a lot more, and spend less money if you use Processing & OpenCV or any of the Processing computer-vision libraries to implement this.

RoboRealm will also do the trick, in fact it will probably be easier, but you will only learn how to use a commercial software which (to my best knowledge) is not especially relevant in the bigger picture.



edit: in case you are not familiar with processing - www.processing.org (or something like that.)

Go Up