[Red vs Blue] [Challenges] [Technology]

Challenge 5 - Ball-grabber challenge

"Tennis ball pickup."

Rules

The initial build regulations would be thus:

  • All parts must be lego.

    No other build rules were deemed neccessary.

    The challenge:

    Eight standard tennis balls will be placed in the centre of the room. The robots will deploy from opposite ends of the room.

    The goal is to pick up the balls and deposit them into one of two boxes. The sides of the boxes are two inches high, and of negligable thickness.

    Each robot must pick up and place two balls in each of the boxes.

    Robots must only pick up a single ball at a time.

    Red Team's Design

    Our initial concepts were ambitious and awesome, but sadly each prototype failed miserably.

    Concept one - automatic grabber.

    If this had worked, our robot would have a massive advantage. A trigger on the front of the robot would engage the grabber motor, allowing us to simply drive full-speed into a tennis ball.

    Result: We were unable to implement a grabber mechanism with sufficient speed to stop the ball bouncing away. Abandoned.





    Concept two - scoop.

    Again, we hoped for an automatic pickup, simply driving at the ball and letting it roll into a holding area. This holding area would then lift and deposit the ball into the boxes.

    Result: The combination of speed and a suitable flat scoop eluded us, as did a reliable lifting mechanism. Abandoned.



    Concept three - super-wide grabber.

    In order to minimise the need for driving accuracy we came up with the idea of a super-wide articulated grabber that would close like a grasping hand. The ball could be grabbed from a non-central position, eliminating the need for careful steering.

    Result: Annoyingly, we could figure out no way to grab only a single ball. If the balls were grouped together, the grabber would try to grab them all at once. Far from ideal. The wide grabber also proved to be a massive hindrance, snagging on obstacles. Abandoned.

    Final Design:

    Our final design began with a pair of wheel modules, each a single heavy wheel with a non-grip slippy wheel for support at the back. When assembled either side of our robot we would have a skid-steer system. The battery box and IR pickups would be placed centrally at the rear of the chassis, and the grabber mechanism would sit between the primary wheels.

    Our grabber claw hangs vertically. Driving the main wheels either side of the tennis ball allows the grabber to close from behind and in front, lifting the ball from the ground.

    The grabber is fixed to a vertical arm which can swing around a central pivot point, raising the ball up and forward for dropping into the boxes.

    The Event!

    Red team's robot
    		 Blue team's robot

    Blue team's machine was a simple, yet impressive, piece of engineering. Their design succeeded in doing what we had failed to manage, in eliminating the need to accurate control.

    The conveyor system scoops up any ball the robot drives into, ready to be launched from the rear of the machine. Genius.

    We set up the boxes and decided on a 'best out of three' trial.

    First run: 4 to 1 Victory - Blue Team

    Second run: 4 to 2 Victory - Blue Team

    There was no need for a third run.

    Post-Challenge Analysis

    The blue team machine was better suited for the task, without question. Our design, while reasonably good at the job, was let down in several areas:

  • Traction. In spite of the giant tyres, the laminated flooring used for the challenge was far slippier than we'd expected. A slight difference in balance rendered one of our wheels useless, leaving our robot unable to turn effectively. Blue Team built their machine on-site and were familiar with the surface.
  • Signal loss. For some reason our tranmitters weren't working properly. The robot was unresponsive, halting or continuing actions for seconds after the controls were released. Annoying.
  • Bizarre misleading tactics. Not for the first time, (the third, I think), blue team had showed us previews prior to the challenge and then produced a robot completely dissimilar. The inital challenge brief was also worded in a way that suggested a design other than the final machine. Odd behaviour, though in this case they let slip a useful idea that helped our design.

    All whinging aside, the conveyor-belt design was far superior. Well done to the Blue team.

    I still think ours looks better