Feb. 9 2018
Created team
Feb. 17 2018
Brainstorming Session
Game Idea List
-motor forward movement & stopping
-line sensing
-motor turning movement
-self adjusting for line following
-mechanics to deposit balls
-tilting system using string & motor
-detect frequencies emitted from beacon
-power switch
-backward movement to refill with balls
Robot Idea List
Created team
Feb. 17 2018
Brainstorming Session
Game Idea List
- Only score in Round A —> during competition only
- Opponent has to outscore you, and you have a head start
- Opponent can't come across bridge until they beat you
- Simple
- Single task — efficiency maximize
- Put buzzword in funding office A
- Put buzzword in funding office B
- Go to brick’s garage
- All of this within 2 minutes 10 seconds
-motor forward movement & stopping
-line sensing
-motor turning movement
-self adjusting for line following
-mechanics to deposit balls
-tilting system using string & motor
-detect frequencies emitted from beacon
-power switch
-backward movement to refill with balls
Robot Idea List
- Simplicity Dump Bot
- Bucket dump
- Bucket holds four balls in a horizontal line
- Dumps all at once
- Extending arm?
- Could get pretty intense
- Make an arm that gets longer and basically just lowers to score in either zone
- Would need the robot to be fairly tip resistant in this case
- Single long arm?
- Can it reach the drop point from the line?
- Should it fold back over or just start up and swing down?
- What is the best orientation for speedy scoring?
- If we make it inline with the drive, then can we us the sensrs accurately enough to guarantee the speedy forward dump
- Should we just make it sideways? So the robot only has to sense the line to make the dump?
- Super lightweight
- Literally the most minimalist frame ever
- Like a single board probably around 100 square inches
- Maybe final design can have it cut from some thin aluminum
- Super fast
- Probably want to use only dc brushed motors for drive
- Gear them to optimal speed with estimated weight
- Maybe embed encoder in transmission?
- Loading mechanism
- Extra thing to put balls in which transfers them to the robot with greater speed and precision to make loading easier
- Tilting mechanism
- Use an arm to push down and tilt the funding rounds in our favor
- Bucket dump
- Drive and Shoot Bot
- Load and shoot all 4 balls at once.
- Probably via a catapult or something
- Maybe spring powered launcher? Like pinball machine launcher
- One calibrated shot which can be repeated consistently
- Probably via a catapult or something
- Idea is to be able to shoot the close goal A from just outside the load zone and then be able to shoot the far goal B from the line in front of A
- This should allow for a greater number of cycles than the dump bot
- It is also less reliable… or more of a challenge to make reliable
- Robot should never have to veer off the line (always use the line as an aiming tool)
- Load and shoot all 4 balls at once.
- Super Shoot Bot
- Shoot all goals with precision from just outside the loading zone
- Super accurate shooter
- Single ball at a time
- Definitely would need to experiment
- Hammer
- Linear Catapult
- Traditional Catapult
- Double wheel
- Single wheel
- Pneumatic mechanism
- Electro Slider
- Lots of experimentation with sensors
- We would probably want some kind of a turret system which rotates on top of the robot… maybe
- Aiming is possible with the drivetrain but it is generally less reliable than a separate aiming system just for the shooter
- If we could get accurate enough with the drivetrain, it would keep cost and weight down
- Aiming is possible with the drivetrain but it is generally less reliable than a separate aiming system just for the shooter
- Maybe we could built some sort of tunnel vision thing around a sensor to make the tower sensing more accurate
- We could have multiple sensors next to each other with increasing tunnel vision to make it really easy to hone in
- What kind of sensor?
- We would probably want some kind of a turret system which rotates on top of the robot… maybe
- Arm+slide robot
- Arm loads the buzzword like a tractor
- Arm rotates past vertical to drop buzzword into a slide/funnel that directs it
- Could potentially load up to 4 buzzwords and then direct all of them
- Plus side-simple with direct loading strategy
- Down side- not as fast as catapulting ideas
|
Feb. 18 2018
|
|
Feb. 19 2018
We decided to go with a simple ball dumping robot. It will have 2 slides that will be actuated separately. The first slide will dump the buzzwords it holds into the first bin, and the second slide will dump buzzwords into the second bin.
We also decided to build the drive train from Legos to allow for easy interface with the Lego wheels
Printed a motor to Lego shaft coupler
Assembled Lego transmissions
Note: The coupler is not perfectly linear, so we are using a jointed axle to allow for some flexibility
We decided to go with a simple ball dumping robot. It will have 2 slides that will be actuated separately. The first slide will dump the buzzwords it holds into the first bin, and the second slide will dump buzzwords into the second bin.
We also decided to build the drive train from Legos to allow for easy interface with the Lego wheels
Printed a motor to Lego shaft coupler
Assembled Lego transmissions
Note: The coupler is not perfectly linear, so we are using a jointed axle to allow for some flexibility
|
|
Feb. 20 2018
Tests with 5:1 transmission demonstrate a need for higher torque
Rebuilt transmission with 25:1 ratio which also allowed us to move the motors to the same side as the wheel for extra lateral space
Connected transmissions together and added swivel wheels to the front
Laser cut base plate to attach to Lego structure
|
|
Feb. 21 2018
Tested light sensors and added them to the base design
Calibrated the sensors to determine black, gray, green, and white colors
Feb. 22 2018
Created a PD controller for line following
Feb. 24 2018
Tested and calibrated line following
The motors still don't output enough torque to reliably move the robot, so we added the second battery
Prototype the scoring mechanism with cardboard and toothpicks
Tested light sensors and added them to the base design
Calibrated the sensors to determine black, gray, green, and white colors
Feb. 22 2018
Created a PD controller for line following
Feb. 24 2018
Tested and calibrated line following
The motors still don't output enough torque to reliably move the robot, so we added the second battery
Prototype the scoring mechanism with cardboard and toothpicks
Feb. 26 2018
Note: At 14V, the motors do move the drive train but it is evident they will struggle with the addition of much more weight. For this reason, and because we are worried about the durability of the Lego frame, we have decided to move forward with an almost entirely foam core construction
Designed the upper frame of the robot
Brainstormed additional ideas for scoring chutes
Decided on a rectangular chute construction with a trigger release mechanism. The chutes use rubber bands to be spring loaded. A wire holds them down until the stepper motor turns to release. The stepper motor has two arms that are offset by about 30 degrees so one side is actuated before the other
Hooked up the stepper motor and tested a simple code to deposit the balls
Note: At 14V, the motors do move the drive train but it is evident they will struggle with the addition of much more weight. For this reason, and because we are worried about the durability of the Lego frame, we have decided to move forward with an almost entirely foam core construction
Designed the upper frame of the robot
Brainstormed additional ideas for scoring chutes
Decided on a rectangular chute construction with a trigger release mechanism. The chutes use rubber bands to be spring loaded. A wire holds them down until the stepper motor turns to release. The stepper motor has two arms that are offset by about 30 degrees so one side is actuated before the other
Hooked up the stepper motor and tested a simple code to deposit the balls
|
|
|
Feb. 28 2018
Tested line following with the entire structure on top
Made the robot stop at the first and third gray line. The robot releases a chute at each
Tuned what angle to start the stepper motor at and how much to turn the motor for each ball dump
Mar. 1 2018
There is a very thin black line between field tiles that the robot keeps reading as a gray line. We implemented a memory function to solve this issue. The robot now averages the last 60 values it has read and uses the average to detect gray lines.
Added a turn state to the code to make it around the 90 degree bend that leads to the opponent garage. We played around a bit with how to make the turn and be situated to follow the next line. Our current strategy is to run the left wheel forward and the right back at equal power until the right line sensor reads the line. This seems to be working.
Mar. 2 2018
The turn is not consistent, so we played around with the power values. Now the left motor is receiving about 15% more power than the right. This seemed to help a lot, though we are concerned it will change if we recharge the battery so we decided to not recharge the battery until we are checked off.
The front swivel wheels have been having some problems. We think it is because there is so much weight on them and they are made of Legos. We experimented a bit with other designs and have found one which works
Added a rest sate which the robot now enters when it reaches the opponents garage
First successful full run through, though it still isn't reading the gray lines reliably and is now sometimes not line following well either
Mar. 3 2018
The line sensors are doing an odd thing where they read values in the 100's until they reach a black line. At this point, the values shoot up to somewhere in the 800's. When it comes off of the line, it then sometimes reads values in the 700's or the 100's again. Gray lines can either read values in the 200's or the 800's.
We thought it could be the sun, so we added a shade, but the problem persists
No matter what range the values are in, they still jump when the robot passes a line, so we tried to implement an adaptive reader which compares an average of the last 30 readings to an average of the last 90 to detect jumps, but this is super unreliable.
Mar. 5 2018
CHECKED OFF!!!!!!!!
The sensors magically began working. They still read values in the 100's at the start, but once they reach the 800's range, they stay there. We reverted to the simplest version of the code and expanded the hysteresis range. To exit the hysteresis cycle, it now must go 95% of the way towards gray if it was on white or 95% of the way towards white if it was on gray. Now it seems to work almost every time.
Mar. 9 2018
Competition Day
We decided to try to be competitive. This meant, however, that we needed some sort of mechanism which we could use to tip the goals.
Design Elements:
Game Strategy:
Tested line following with the entire structure on top
Made the robot stop at the first and third gray line. The robot releases a chute at each
Tuned what angle to start the stepper motor at and how much to turn the motor for each ball dump
Mar. 1 2018
There is a very thin black line between field tiles that the robot keeps reading as a gray line. We implemented a memory function to solve this issue. The robot now averages the last 60 values it has read and uses the average to detect gray lines.
Added a turn state to the code to make it around the 90 degree bend that leads to the opponent garage. We played around a bit with how to make the turn and be situated to follow the next line. Our current strategy is to run the left wheel forward and the right back at equal power until the right line sensor reads the line. This seems to be working.
Mar. 2 2018
The turn is not consistent, so we played around with the power values. Now the left motor is receiving about 15% more power than the right. This seemed to help a lot, though we are concerned it will change if we recharge the battery so we decided to not recharge the battery until we are checked off.
The front swivel wheels have been having some problems. We think it is because there is so much weight on them and they are made of Legos. We experimented a bit with other designs and have found one which works
Added a rest sate which the robot now enters when it reaches the opponents garage
First successful full run through, though it still isn't reading the gray lines reliably and is now sometimes not line following well either
Mar. 3 2018
The line sensors are doing an odd thing where they read values in the 100's until they reach a black line. At this point, the values shoot up to somewhere in the 800's. When it comes off of the line, it then sometimes reads values in the 700's or the 100's again. Gray lines can either read values in the 200's or the 800's.
We thought it could be the sun, so we added a shade, but the problem persists
No matter what range the values are in, they still jump when the robot passes a line, so we tried to implement an adaptive reader which compares an average of the last 30 readings to an average of the last 90 to detect jumps, but this is super unreliable.
Mar. 5 2018
CHECKED OFF!!!!!!!!
The sensors magically began working. They still read values in the 100's at the start, but once they reach the 800's range, they stay there. We reverted to the simplest version of the code and expanded the hysteresis range. To exit the hysteresis cycle, it now must go 95% of the way towards gray if it was on white or 95% of the way towards white if it was on gray. Now it seems to work almost every time.
Mar. 9 2018
Competition Day
We decided to try to be competitive. This meant, however, that we needed some sort of mechanism which we could use to tip the goals.
Design Elements:
- Tipper must be very large to expand our effective hit zone
- Must be strong enough to lift at least one side of our robot off the ground
- Must actuate using existing stepper motor in exactly the position it is in
- Must reach the goal while still allowing the robot to maintain its position on the line
- Must be built entirely from foam core
- Must have the capacity to score the preloaded balls
- Must be simple and consistent
Game Strategy:
- Score 4 in round A
- Defend A until 1:45
- Move to and tip B
- Quarterfinalists!! Woo!!
- Score 4 in A 100% Consistency
- Stole a total of 2 balls
|
|
|