6.jpg

VEX Robot

For the 2017-2018 VEX Robotics Competition
Mid 2017

At the beginning of grade 11, my mother suggested me to participate the Robotics Competition. Yet, I had no idea that later I would be a leader who not only leads his own team to victory but also coaches other teams.

 
4.jpg

Prologue

In December 2016, I participated my first robotics competition as the programmer. My team won the third-place, which is a good start. Later, I promoted to the captain of our school team 97917 because of my outstanding performance. After the 2016-2017 Vex season, I took a step beyond and created my own team 1690X for the 2017-2018 Vex season. This was when everything (and every struggle) began.

 

Introducing the 2017-2018 VEX In The Zone

The main task is stacking cones on stationary or mobile goals and delivering the mobile goals to scoring zones.

e33a1584-a_edited.jpg
 
5.jpg

Team 1690X

Bell Chen and I are the founders of team 1690X. At first, my duty is mechanical designing and building. Bell Chen is an experienced programmer. Yicheng Wei is the driver and Xupei Hu is my building assistant and a sundry handler.

 

Designing thinking

 
maxresdefault.jpg

A bold plan

The mobile goals worth much more than cones. As a person who does not like to conform to others, I definitely had a different playing strategy - tip over opponents’ mobile goals so that they cannot score. However, the actual test on the center of mass of the mobile goals showed that even if the mobile goal is tipped over, it could be restored with little effort. This forced me to stop pursuing the first idea.

VEX-Zone-01.png

The focus

Unable to find another shortcut to the competition, I followed the trend of most teams – design a robot that can stack cones efficiently and transport mobile goals stably.

BookScanCenter_6_edited.png

The main arm

To stack more cones than opponents, obviously, a primary arm with a considerable height is required. Although I had some experience in designing the last year’s lifting structure, none of the previous designs was applicable because they were just too short. My teammate Bell Chen suggested me to use the double-parallelogram arm because of its large height after expansion. I later studied this structure and found out that it was really suitable for this year’s game. Therefore, I decided to use the double-parallelogram arm as the main arm.

maxresdefault (1).jpg

The mobile goal arm

The main arm designed for the light cones, apparently, would not be able to lift a heavy object like the mobile goal. To deal with the heavy mobile goals which weigh 1.7kg each, I had to think of another method. Designing a secondary lift only for the mobile goal was the first idea popped up in my mind. Although I really did not like this because it means spending an extra motor on a component that is often not functioning, I could not find a better solution at that time. As a result, I summarized that the basic design must consist of four components – the main arm for cones, a secondary arm for mobile goals, a claw to catch cones and a base.

IMG_20171002_182958_edited.jpg

Stack while moving

Heavily influenced by the idea of “speed is everything”, I started to consider how to increase the efficiency of the robot. Taking cone and carrying it to a stationary goal would be too slow because a lot of time is wasted on running between the cones and mobile goals. Then I considered, how about the robot stacks cones on the mobile goal that it carries? By doing so, the robot could save a huge amount of time on running between the mobile goals and the cones. But how could the robot place the cone on the mobile goal it carries?

13231412.jpg

Two degrees of freedom

It turns out that the arm must have two degrees of freedom to solve this problem. Bell Chen and I watched some Youtube videos. We found that installing a rocker arm at the end of the double-parallelogram is a solution. However, the rocker arm must be independently powered and this leads to an increase in the number of arms. This adds with the other two arms give three arms in total! What a waste of materials and motors! In order to save motor usage and prevent waste, I decided to combine the rocker arm and the mobile goal arm. Then, the claw must also be able to catch the mobile goal.

2-269.jpg

Final plan, motor distribution and sensors

The maximum number of motors that can be used is 12 in VRC so the allocation must be scientific. I decided to use a 6-motor turbo base to ensure speed. For the lift part, I decided to use a 2-motor double-parallelogram main arm with a good power-assisting structure. The rocker arm also takes 2 motors but the gear ratio and power-assisting structure must be adjusted carefully in order to lift the mobile goals. At last, a claw that can catch both mobile goals and cones is required. To make the control more functional, the base has to have encoders and the arms must have encoders or potentiometers.

Early development

4 generations of robot

Before the West China Tournament, I was in charge of designing and building. All the way from 2017/7/3 to 2017/8/15, I was working on the robot for the whole day. To learn about the story and the robot we took to the first tournament, click the link below.

8.jpg
 

Basic introduction to our first competition robot

Our first robot has one base, one mobile goal lift on the base, one claw on the base, one double-parallelogram main arm, one rocker arm on the main arm and one suction structure on the rocker arm, The robot is capable of stacking cones and delivering the mobile goals to the scoring zones. The video shows how it works.

 

Taking the burden

A few days after finishing the 4th generation robot, we participated in our first tournament of the season. Although we strived and finally won the Skill Challenge Champion and Design Award, we failed the final matches and ended up with a First Runner-up.

The Western China Tournament has shown some problems with our robot. The next step is to solve these problems and make our robot better. To avoid the tragedy happening again, every effort must be made to ensure victory.

At the same time, Bell Chen left our team because of his conflict with the A.I.R. club. As a result, the burden of programming fell upon me. 

Now being the team captain, mechanical designer, builder, and programmer, I had so much to carry.

IMG_20170820_161444_edited.jpg
 
IMG_20170824_184512.jpg

Ideas about upgrade

Lesson from the West China Tournament

There are in total three key points: Increase speed, prevent paralysis and change the mobile goal arm.

The purpose of increasing speed, obviously, is to maximize efficiency. The base, mechanical arms, and programs can all be upgraded to achieve this purpose. The weight of the robot must be reduced also.

Preventing paralysis is the most important thing. Did our robot not stick on the pipe on the field, we would not lose the first match in finals. As the chief designer, I had the most responsibility for this, and I must prevent such a ridiculous thing from happening again.

The mobile goal arm cannot directly transport the goal into the goal zone while the robot is out of the goal zone. Therefore, changing it so that it can transport the goal into the goal zone without the robot itself passing the 10pts zone bar is the main idea. This is the most difficult part because it will change nearly all the other structures.

 

The final design

After I became in full control of our robot

01.jpg

The main arm

First of all, in order to make space for the mobile goal lift which might be larger in this case, I moved the main arm more outward. However, I found the arm to be easily deformed.

02.jpg

Solving the rigidity problem

Low rigidity has always been a problem for any double-parallelogram arm, and strengthening channels must be installed to connect the arms at either side. To help design such a strengthening structure, I analyzed the available locations to install the strengthening channels on the arm and later I made several modifications on the arm according to the result of the analysis.

03.jpg

Strengthening structure

The horizontal channels are the new strengthening structure.

04.jpg

Farewell to the Xixi kun

An additional claw must be installed because of the Xixi Kun. Therefore, if this design is given up, at least 1kg of mass could be reduced. Considering that a good claw is a lot faster than the Xixi Kun and the excessive motor can be added to the rocker arm to increase speed, I decided to say farewell to Xixi Kun which is once the trademark of our robot. It pained me greatly, but to ensure efficiency, I must exclude my personal feeling when designing.

05.jpg

Claw opening and its interference with the rocker arm

However, the maximum opening of the claw is limited by the spacing of the rocker arm which in our case is only 6.5 inches. Since such opening is too small to catch a cone which has a base diameter of 6 inches, I decided to restrict the maximum angle of rotation of the rocker arm and increase the height of the claw to avoid it entering the spacing of the rocker arm so that it does not need to be smaller than 6.5 inches.

06.jpg

Rocker arm change

Next is the rocker arm. To add the other motor, I tried some different turbo motor (240RPM) 5:1 designs. However, none of them seems to work. A thought then popped up in my mind – what is about using the 36T gear connected to a standard motor (100RPM) to power the 60T gear directly. Calculations showed that this would be 25% faster than turbo 5:1 theoretically, which is exactly what I am looking for! Then I changed the speed configuration into standard motor 5:3 and produced a draft.

09.jpg

The arms and the claw

This is the main arm, rocker arm and the claw all put together.

10.jpg

Oversizing

However, I found that the rocker arm must be shortened because it not only goes beyond the size limit but also worsens the deviation range of the center of mass of the robot. However, this would make the stacked cones move into the robot that increases the possibility for them to interfere with the lift. The spacing between the output shafts of the current claw is too small so the cone will move further after being caught. The solution is to increase the spacing by changing the claw into a standard speed 3:1 one.

11.jpg

The hat claw

The claw was then modified. The hat is added to making it easier to deal with preload cones, the effect will be later shown in the video.

12.jpg

The mobile goal arm

The next challenge is the mobile goal arm. There are two goals of design. First, the robot must be able to put the mobile goal in the 10pts zone without the riding over the bar. Second, the goal must not slide out. After measurement on the field, the AB was chosen to be 6.5-inch long. This could avoid interference with the arm while achieving the above aims.

14.jpg

The actual mobile goal arm

This is the complete mobile goal arm.

15.jpg

Finished?

Starting from September, in total I spent 1 month getting the robot look like this. However, there are still some problems with the details. Throughout October, I was focusing on improving the robot by frequently making small modifications.

 

At long last

You cannot imagine how much effort I have put on my robot. After a total of 3 months of hard-work, with 5 unsuccessful prototypes being made, I have completed the final version of our robot in November 2017. This is the final robot that I designed for the 2017-2018 Vex Robotics Competition. Overall, it has 1 base, 3 arms and 1 claw with a hat. It weighs only 8kg and was once the most efficient VEX robot in the world.

16.jpg
 
Engineering Plans

The final presentation

The fragrance of plum blossom sharpens in the bitter cold

 

The base

The base has 6 motors powering the 4 4-inch Omni-wheels with direct-drive. The motors were in turbo-gear-set (240RPM), making it one of the fastest robots. The motors on each side are linked by gears and sprockets to ensure even load distribution. The base has encoders on both sides so that it can be controlled more precisely during the autonomous period.

IMG_20171224_213549.jpg
IMG_20171224_213549.jpg

press to zoom
IMG_20171023_194436.jpg
IMG_20171023_194436.jpg

press to zoom
IMG_20171112_130325_1.jpg
IMG_20171112_130325_1.jpg

press to zoom
IMG_20171224_213549.jpg
IMG_20171224_213549.jpg

press to zoom

The arms

These are the main arm and the rocker arm. The main arm uses a double-parallelogram mechanism and the rocker arm uses a chain. The main arm is powered by 2 standard motors (100RPM) with a 5:1 external gear ratio. The rocker arm is powered by 2 standard motors (100RPM) with a 5:3 external gear ratio. Working together, they can stack 17 cones in total.

IMG_20171223_173729.jpg
IMG_20171223_173729.jpg

press to zoom
IMG_20171223_174014.jpg
IMG_20171223_174014.jpg

press to zoom
IMG_20171223_174554.jpg
IMG_20171223_174554.jpg

press to zoom
IMG_20171223_173729.jpg
IMG_20171223_173729.jpg

press to zoom

The claw

This is the claw. It is powered by 1 high-speed motor (160RPM) with a 3:1 external gear ratio. It has a hat that makes stacking preload much easier. The zip ties on the inner surface of the claw help it to catch cones very securely.

IMG_20171112_130057_1.jpg
IMG_20171112_130057_1.jpg

press to zoom
IMG_20171112_130100.jpg
IMG_20171112_130100.jpg

press to zoom
IMG_20171112_130738.jpg
IMG_20171112_130738.jpg

press to zoom
IMG_20171112_130057_1.jpg
IMG_20171112_130057_1.jpg

press to zoom

The mobile goal arm

This is the mobile goal arm. It is powered by one standard motor (100RPM) with a 9:1 external gear ratio. It holds the mobile goal securely on the robot and it can perfectly deliver the mobile goal to both the 10pts zone and the 20pts zone.

IMG_20171223_173844.jpg
IMG_20171223_173844.jpg

press to zoom
IMG_20171224_162647.jpg
IMG_20171224_162647.jpg

press to zoom
IMG_20171223_175048.jpg
IMG_20171223_175048.jpg

press to zoom
IMG_20171223_173844.jpg
IMG_20171223_173844.jpg

press to zoom

​The motors

The mechanical properties of motors vary. Therefore, motors must be carefully selected. I have designed a motor tester in which the motor being tested will drag an unplugged motor (this provides a constant load) and the potential difference of the unplugged motor as shown by the voltmeter indicates the rotational speed of the motor being tested. Once the data is recorded, by comparison, we can select the best motors.

IMG_20171224_173438.jpg
IMG_20171224_173438.jpg

press to zoom
2-269.jpg
2-269.jpg

press to zoom
IMG_20171224_173438.jpg
IMG_20171224_173438.jpg

press to zoom
 

The code

The robot uses both closed-loop and open-loop controls. It has a manual control program with overload protection. It has 5 sets of PID for the autonomous period and has a semi-autonomous stacking program for the driver-control period. According to the driver, the robot is very easy to operate.

3241.jpg
 

The Power Assisting Structure

The brightest spot of the robot is the power assisting structure. I have done research on connecting rubber bands and summarized it into a book. With the power assisting structure, the torque from the weight of the lift at any expansion level is offset, which drastically lowers the power needed for the lift structure, and in fact that primary lift uses only two 393 motors, which each has a maximum power of only 3W. For more information about the Power Assisting Structure, please click the link below.

IMG_20171223_175017.jpg
 
psbON8NQT1B.jpg

Our show

At the time I finished my robot, 1690 was no longer fighting alone. Many other students joined us and I helped them to create their own teams. 1690Y, 1690Z and 1690K. I shared a lot of my experience with the teams and helped them with designing, building, and coding. I was no longer just a team captain, I was a coach. And at last, I have led them to victory.

For the 2017-2018 season, we have won 1 tournament champion, 2 robot skills champions, 2 tournament second-places, 1 tournament third-place, 1 design award, 1 think award in 4 tournaments. Our robot once made the national high school record for the skills challenge.

 
 
 
P70816-200740_edited.jpg

Epilogue

1 year 2 months and 24 days, participating in the VEX Robotics Competition is absolutely the most painful but most meaningful thing I have done in high school. Its terribly commercialized environment, its low degree of freedom and its ridiculous cost made me suffered. Nevertheless, it not only made me learned and applied a lot of engineering knowledge and improved me tremendously in engineering skills but also established my central ideas of design: efficient, light and as simple as possible. Most importantly, it made me a leader who is not only able to improve himself but also able to influence others positively.

At last, I must say thanks to all the organizations and all the people that helped me during this precious journey.