Вы находитесь на странице: 1из 14

Marble Sorter 3.3.

1
Miles Simpkins
Group: Reid Harrison, Justin Alpern, Ariel
Whitehead
POE Block 3
Westwood High School
3/24/16
Design Problem
The National Recreation Park Association (NRPA) has placed recycling bins
within their parks and constructed a sorting facility. The sorting facility needs a
mechanized system that can sort the different recyclable materials.
Our task is to design, model and test a system that sorts the recyclable materials
given these constraints:
The process must be fully automated.

Only the use of teacher-approved materials or items VEX kit are


approved.
The system must separate mixed marbles into separate holding bins.
Marbles must be fully controlled throughout the process.
16 total marbles of 4 different materials must be sorted. For extra credit,
20 must be sorted of 5 different materials.
The whole process must be finished in under 2 minutes.

Brainstorm Idea
The four marbles I chose to sort are steel, clear white, and aluminum. They begin
in a bin or hopper with a track that leads down to a servo powered door that feeds them
onto a track one at a time.
The first station is an electro magnet that will capture the steel ball and drop it
into a bin.
The next station is a light sensor and flashlight that will sort out the clear marble
(by detecting light shining through) and a solenoid will punch it onto a separate track
down to the clear bin.
The final station is under a canopy and uses a light sensor or line follower to sort
between the white and aluminum marble. After the light sensor recognizes each marble,
a servo arm knocks the marbles one way or another down to their respective bins.

Decision Matrix
Group
Members

Difficulty of
Programming

Construction Speed
Time
of
sorting

Accuracy of
sorting

Requirements
met

Total

Ariel

18

Reid

18

Justin

18

Miles

19

1 --------------------- 5
Worst
Best

We began the design making process by having each team member present
their design proposal. Afterwards the team assigned scores based on difficulty of

programming, construction time, speed of sorting, accuracy and whether or not it met
the requirements of the task.
Accuracy was based on how well each marble sorter could tell the difference
between marbles. For example having one light sensor decipher between all four
marbles would score low in accuracy. This criterion was important for the success of the
sorter.
Difficulty of programming was scored on how complex each program was
predicted to be. Factors like timing, the number of loops required, etc determined a
program's complexity. This criterion was important because of the time constraint our
team was under.
Construction Time was the predicted time for hardware to be constructed based
on its complexity. A sorter that was too complex mechanically would take too long to
build and calibrate properly. Time was very important, and our team was working under
a tight schedule.
The Speed of Sorting criterion was important to meet the requirement of the
design brief of two minutes. This was also determined theoretically based on how many
marbles could be sorted at once, accurately.
The Requirements met criterion was necessary to ensure each machine would
complete the task. This was assessed by checking the design problem and making sure
each constraint was met.
Our group selected Miles design because because it was accurate and had the
potential for quicker sorting. It incorporated effective ideas that other designs already
had. For example, Justins use of a servo trap door at the beginning of the sorter was
already incorporated into Miles design.

Final Design Solution

We chose my design mainly because of the accuracy it has having a station


cater to each marbles needs. It also allowed for stations to be modified as we went
along. My design was chosen without modification.
Design Modifications
Throughout this project our team made numerous modifications to the hardware
and program of our marble sorter.
The first modification occurred early in the building and programming process
when we realized that the electromagnet was not powerful enough to attract the steel
marble, the electromagnet being offset from the track. To solve this, Reid built what we
called Reids Bar which was an extendable arm powered by a motor and guided by a
gear and gear track. The electromagnet was then fastened to the end of the bar,
allowing it to come into direct contact with the marbles.
Our second modification was replacing the solenoid after the LED/light sensor
station with a servo powered arm. This modification was necessary because the
solenoids reach was not far enough to knock a marble resting in the track off the other

side into a bin. To solve this Justin assembled a servo swinging arm to clear out a track
compartment if deemed necessary.
Our third modification was replacing the servo door with a solenoid. We made
this modification because our two other servos were now being used elsewhere on the
sorter. To add, using a solenoid that could quickly extend and retract probably turned
out to be more effective than a servo door.
Another hardware modification resulted from an issue we had dropping one
marble into its designated track compartment consistently. This was due to the marbles
having different weights based on their materials. Some would roll faster than others
and overshoot the designated compartment. To solve this the hardware team installed a
backboard to deflect any overshooting marbles back down into their proper
compartments.
Our next modification was to improve the consistency of advancing the belt. This
modification was done prior to any major testing and involved adding an encoder to
track the revolutions of the wheel powering the belt, theoretically improving accuracy.
In another modification, after testing the belt advancement numerous times and
finding it to be inconsistent, we attempted to decrease the belt drift after turning off the
motor which powered the belt. Belt drift refers to the momentum of the motors carrying
the belt too far forward after being powered off. These drifts were inconsistent due to a
lot of friction caused by the belt rubbing into components of the marble sorter, and
consequently jerking forward to overcome obstacles. To solve this problem we created a
gear train with a high gear ratio, meaning a lot of turns of the smaller, powered gear
would result in fewer revolutions of the gear attached to the same axle as the belt. This
modification reduced the belts tendency to drift.
Our final modification was a programming one, to reduce the belts inconsistent
movement. The initial program required the belt to move in reverse if a clear or
aluminum marble was detected, to return the belt and other marbles on the track to their
respective stations. This would ensure that no marble skipped a station. We found that
whenever the belt reversed its direction, the station alignment with the compartments
was thrown off. Unable to find the source of this problem, we modified the code to run

only one marble through all stations before releasing the next one. This dramatically
increased our sorting time but was necessary to successfully sort marbles.
Final Design

Our machine functioned as it was expected to function, having three of the four
marbles making it through the sorter. Only four marbles were attempted to be sorted,
the new code clearly slowing down the sorting process. Three were sorted into separate
bins, and the last one fell off of the track. Unfortunately we were unable to calibrate the
LED/light sensor and servo station to properly align with the belt. Due to the same issue
of belt consistency were unable to test the line follower and servo station.

ROBOTC Program

Design Process
Define the Problem: We began the marble sorter project by being presented with a
design problem from NRPA with the criteria and constraints. Before moving on we
ensured that we understood the problem NRPA had in sorting recyclables and
recognized the constraints, therefore identifying the problem.
Generate Concepts: The next step was individually brainstorming ideas by sketching
and planning marble sorters. Once each team member was finished with their design,
we agreed on decision matrix criteria and each member presented their solutions to the
group. We then evaluated each proposal and selected the best one.

Develop a Solution: Our next step was for each team member to sketch the selected
design, creating technical drawings to ensure they had the knowledge on how it worked
and begin planning on how to construct hardware and the program. It was designated
here that Ariel and Justin were the hardware team, and Miles and Reid would program,
Reid helping out with hardware when necessary.
Construct and Test Prototype: After technical drawings and a plan were acquired, our
team began constructing hardware and writing code. We tried to test hardware with
code as we went along but this proved to be tough due to the different paces of the
teams. As we tested, we analyzed each issue we ran into and created design
modification to solve problems.
Evaluate Solution: After each design modification we ensured each one solved the
given problem. Also Mr. Landers was conducting test reports which made each team
evaluate their current position. Mr. Landers and his team (Mr.Dipalma) also provided
project recommendations.
The construct and Test Prototype and Evaluate Solution steps went hand in hand and
numerous times we jumped between the two.
Present the Solution: At the end of the project we presented our solutions to the class,
and tested them for effectiveness. This presentation also included the problems we
faced and how we overcame them.
Team Evaluation
Reid Harrison showed by far the most dedication to the marble sorter project by
being a part of the hardware and software team, and coming in the most outside of
class with both the hardware and software teams. Reid was polite and never missed a
day. Reid therefore, upheld the group norms and went well out of his way to help the
team.

Ariel Whitehead as part of the hardware team, fulfilled the group norms. She was
very polite and came in outside of class on numerous occasions. She stayed on task,
and let the group know ahead of time on her only absence during the brainstorming
phase. She even went out of her way by writing the team a note and bookmarking her
brainstorm idea and description in her engineer's notebook allowing the brainstorming
phase to go smoothly without her.
Justin Alpern was part of the hardware team, and upheld the group norms by
completing his work on hardware. He remained on task and even came in outside of
class to work on hardware with the hardware team. He developed good solutions to
problems like the servo powered arm.
I was on the programming team, and wrote the majority of the code along with
Reids recommendations and troubleshooting. I came in numerous times outside of
class to work on calibrating the code with hardware and attempting to troubleshoot our
belt issues. Therefore I upheld our group norms.
Reflection
a.) My team would not rely on the conveyor belt to transport the marbles to each station.
The entire process depended on the ability of the conveyor belt to move marbles
consistent distances. We were never able to figure out how to achieve this, and in turn
our marble sorter failed.
b.) The most challenging aspect of this design problem was sorting four different
materials. Our machine could sort the steel marble often and the clear marble rarely.
Our last station was never calibrated. Therefore we had a sorter that could only sort two
of the four marble materials.
c.) The most important thing I learned is to integrate different aspects of a project as
quickly as possible for testing and evaluation. Our group made the mistake of not
applying programming to hardware until fairly late in the construct and test a prototype
phase. As a result, we had less time to troubleshoot. Since we relied so heavily on the
conveyor belt, it would have been smart to test that as quickly as possible. Because if

the belt consistency problem was caught early on, we may have been able to change
designs. Also, I learned that the encoder reads best when the shaft it is reading from is
attached to a gear resting on the encoder, which provides consistent motion. Also I
learned a life lesson. In doing this experiment and failing I learned more engineering
strategies than if I had succeeded.
d.) There were a few challenges that resulted from working in a design team. The first
challenge was having everyone on the same page. A few times during hardware
construction, components were built with the wrong specifications. For example, the
ramp at the beginning of the design was built way too steep. My design called for a
gradual slope down. Another challenge that became apparent early on was the faster
pace of the programming teams progress compared to the hardware. I found myself
waiting and helping out with the hardware, my program unable to be improved without
testing until the hardware team was ready.

Вам также может понравиться