Week 7

PROGRESS

Our main accomplishments this previous week are as follows:

Manufactured End Effector Mounting Adapter
Manufactured and assembled robot chassis
Made first version of firing servo mount
Received and assembled PCB
Began firmware for microcontroller

End Effector

We manufactured our design based on our CAD from the previous week. We aim to have this whole system finalized by the end of the day on the 18th (tomorrow). Most of the work in mounting the end effector to the arm is complete, now we just need to actually do the final attachment, and worry about tube and wire routing.

Robot Chassis

We implemented the design from our CAD that we presented the previous week. All of our 80/20 was cut to length and assembled per the design. It works well, and provides us with a very sturdy base for our robot. We have realized, after manufacture, that it still needs another iteration, as the top of the robot isn't low enough for us to be able to do the movements with the arm that we need to do. Luckily, there is sufficient room for the robot to lower the robot between 6" and 12". The one final thing we need to attach is the locking casters so that we can move the robot around.

FiRING SERVO MOUNT V1

We made the first version of our firing mechanism this week, however, it definitely needs another version. This is the mechanism that releases the pressure in our firing mechanism automatically when the robot is ready to fire. It's a pretty simple 3D printed assembly, and as such, will be easy to iterate on.

DENSO Support PCB

We received and assembled our support PCB. It's essentially just an Arduino Shield that contains all of the custom electronics that are necessary for our setup (except for our mains to 24V DC converter. This shield allows us to make nice cabling for all of our electronics in a permanent, robust wiring harness.

FIRMWARE

Firmware for the SparkFun Redboard that is being used as our main embedded MCU has also been started. We've begun writring our own custom interface on the microcontroller side that will enable Command & Control from our computer. We have made significant progress on this code, as the interface is almost entirely done. It can also control our firing servo, read from our pressure sensor, control indicatory LEDs, and charge the pressure vessel (at least logically). We still need to write a ROS node to control this from the computer.

The rest of our time was spent working on embedded firmware for the uno to get control of our firing servo, reading of the pressure sensor, and charging of the pressure vessel. It also includes indication to the user via led strips.

Ball Tracking

Over the last week, we continued work on detecting the ball with ROS, though unsuccessfully. We then transitioned into playing with TensorFlow to do human and ball detection in a 2d plane, however, Corwin only got so far as installing tensorflow and running a test program.

Challenges

A few of our challenges this week:

Creating an arc for the end effector was hard - had to be done by hand (see picture to the right)
Need another version of 3D printed firing assembly
We didn't choose a powerful enough linear regulator to power our microcontroller and all of the LEDs that we want to implement (probably)
Serious issues with ball detection - if needed we can design a simpler catching system

Plan for Next Week

During the next week we will:

Attach the end effector to the arm (by tomorrow)
Iterate on the robot chassis
Attach casters to the robot
Finish wiring up all our electronics
Begin serious software testing and integration
Decide on a final solution for what we want to do with CV.
Hopefully play a practice round

Google Sites

Report abuse