initial idea :

conveyor types :

Before beginning to design anything, I realized there was a need to clarify what kind of conveyor I would be making, as there are various types.

I was able to identify three types I was potentially interested in making :

1 ) the roller conveyor : rolling dowels in continuous rotation allow objects to move across their surface
2 ) conveyor belt : two rollers in continuous rotation allow a belt to move across them
3 ) plate conveyor : plates are pushed via short vertical dowels in oppositional rotation *diagram 3 is in plan*

julio's conveyor belt : (source : hbo)

As previously stated, Julio Torres' HBO Special, "My Favorite Shapes" was my initial inspiration.As such, I decided to look at what kind of conveyor his was. His conveyor curves from the back of the stage to the front, but it doesn't seem to be one continuous piece, which I am interested in.

close up of julio's conveyor belt: (source : hbo)

What I really like about Julio's conveyor is that even though it has curves, the conveyor reads as continuous throughout. From my initial research, I found that often, when conveyor belts curve, they introduce a different kind of conveyor type, making the conveyor's form read as broken up into various modular pieces

Julio's conveyor belt initially confused me because even though it looks like a Type 2, classic conveyor belts cannot, in their conventional form, take curves. Yet the entirety of Julio's conveyor belt reads as homogenous. It can't be a Type 2!

vulture article : here

After doing some research, I discovered the belt was manufactured by a company that often works for sushi restaurants, which I would say falls into Type 3.

sushi belt's circular / crescent plates : (source : wall street journal)

I didn't initially think Julio's conveyor belt was a plate conveyor belt, and I especially did not link it to sushi conveyor belts because those typically have a recognizable circular plate that allows for plates to take a curve, kind of like a baggage carousel.

Knowing I wanted to make circular, continuous form, I decided to pursue Type 3, though I would try to design a plate that, as in Julio's conveyor belt, did not reveal it's mechanism.

conveyor shape :

Initially, I was interested in both a kind of pill shape and a circular shape. In general, I wanted a looping conveyor belt.

Researching the different types as well as just thinking about what the different profiles provided helped narrow my choice.

shape a : pill

SHAPE A : meant non/continuous curvature. That is, you have two straight segments, and half donut shapes at the ends that connect the two. This reminds me of the very typical sushi conveyor belt.

shape b : circle

SHAPE B : being a circle, meant continuous curvature. In this sense, I wouldn't have to conceive of the plate as taking a curve, rather the plate is curved, though this is curve would be quite subtle. Shape B also begins to become spatial.

circular conveyor belt forms a small "room"

With the correct proportions / dimensions, I could imagine someone being inside this circle, presenting the objects on it (like Julio does in his show), while people walk around it.

belt made of 4 modular parts

Although I want one continuous conveyor belt, due to it's size (hopefully an outer diameter of around 6 feet) it's possible it will have to be fabricated as smaller parts that, then, in their assembly are camouflaged into looking as though it is one continuous circle.

Thus, I am thinking about dividing the circle into 4 quadrants.

one plate :

Now that I've vaguely defined the overall dimension / proportion / shape of my conveyor belt, I think it is time to diagram how one plate would move along the circle.

single curved plate : plan

As mentioned previously, due to the circular shape, the plates will have a slight though continuous bend.

diagrammatic kit of parts :

In order for the plates to move, the belt needs to incorporate the following :

1 ) rotational gear : unsure of how many need to be motorized vs connected to motorized gear *note that gears across one another rotate in opposite directions in order to push plate in and away
2 ) plate : above, hiding mechanism below
3 ) gripping band : connects gears / where surface of moving band begins to move base plate
4 ) base plate : connects plates to one another and allows plate to sit in conveyor belt base
5 ) rail / channel : in which base plate sits

diagrammatic section :

1 ) plate : 5" long / 2 ) base : outer diameter of 6' / 3 ) rail / channel : holds plate / allows movement via gears / 4 ) lower cavity : to store motor / wires / etc / 5 ) motor : currently modelled as "Stepper Motor NEMA-17" / 6 ) gear with bands : upper bands connect pair / lower bands connect neighboring pair

I can't say whether any of this is actually true. Sounds like it might be time to prototype some gears?

simplified section and other ideas : ty paloma !

I asked to meet with Paloma to talk about my final project. I had gotten a quick "yes, go ahead" from Jen and Neil, but I had yet to discuss specifics with anyone.

Paloma's advice was to simplify the moving mechanism. Yes, the conveyor belt is quite large (which she suggested considering making smaller, but I do feel strongly about making it 6' wide) but it can be thought of kind of like.. a toy train that just moves in one circle - in that sense it's not really even turning.

new conceptual section with lip + ball + wheel :

Paloma also mentioned that I would want to incorporate:

1. lip detail : that kept plates from popping out of the belt
2. ball detail : in order to reduce friction, and thus, heat
3. wheel mechanism : that would pull the plates along the circle - though her friend who was at office hours (whose name I missed), suggested this wheel be stationary. Paloma mentioned this wheel could be a rollerblade wheel or I could make my own.

Along these lines, Paloma suggested taking a look at Jackie Berry's final project, which included rubber cast wheels, which was super helpful. Jackie's mini robot also included a adafruit 3/4" metal ball caster, which then pointed me to looking at metal balls for the other detail.

new section drawn :

According to Paloma's feedback, I redrew what could be the section for the conveyor belt. A lip holds the plate in, small 1/2" stainless steel balls allow for the plate to move with little fiction, a stationary wheel is motorized / held by two stepper motors.

The last detail, I'm most unsure about, but I figured two stepper motors supporting one wheel could either 1) be twice as strong, or I could switch motors for different directions, always using one motor as a kind of ball bearing. That is, I am using two motors to motorize the wheel but also to support movement of it - not sure if that makes sense.

*Paloma did not suggest a stepper motor, as stepper motors deal with a kind of precision I don't really need for this project. BUT, alas, I had purchased a couple after seeing the 3D printed motorized sphere in class. I was impressed with the little amount of sound it made - sounds it also someting I am worried about - and how fast it moved (thought I wont need ULTRA speed and I wont be carrying huge items). Admittedly, it was a naive purchase, but I don't think it would be totally inappropriate to use it?...*

10 / 20 / 19 :  no nO No NO stepper motor!

Today I met with Anthony to speak about my board that wouldn't blink, but once we figured that out, we discussed my final project briefly.

some things came up :

1.) I should not use a stepper motor! Anthony said this would complicate my project unecessarily.

2.) Like Paloma, he suggested a DC motor would be appropriate for my project. We both agreed, I would need a somewhat powerful motor in order to move my conveyor belt, which, while thin is big and made of aluminum(fingers crossed).Anthony suggested I ask for a DC motor from RPL and test it beginning with a dc power supply.

3.) it's likely the motor set up would be : motor shaft going through wheel and then into a bearing - load is reduced via the steal bearings, but there is still likely to be some torque on the motor, thus the need for a bearing connected to the end of the shaft turning the wheel.

4.) rather than assume perfect tolerances, the motor / wheel assembly should be adjustable : that is it can move up or down in order to get the moving wheel making contact with the plates so that they moving as they should - a heavier plate might also help...

5.) The plates should be linked or be incredibly thick, in order for the plates to not be pushed on top of one another. This linking mechanism does not have to be aluminum, but it also doesn't have to respond directly to the wheel and steel bearing below it, as previously drawn.

6.) The wheel that will move the plated should be flat, if possible, not rounded as currently drawn.

10 / 21 / 19 :  playing w a dc motor

***a lot of this section is discussed in WEEK 07***

Yesterday, Anthony mentioned I should try to play w a dc motor, see what its like, possibly prototype a scaled version of my conveyor belt with it. That sounded easy enough, what I was intimidated by was making a circuit board that could control a dc motor. So, I attempted both.

Neil had a dc motor board, which I redrew in Eagle :

10 / 22 / 19 :  dc motor arduino code corrected

Today I went to office hours, and Anthony revealed I had selected the incorrect clock and thus my delay was off by a factor of 20!

I though I would have to make a vastly different board in order to power more powerful motors, but Anthony also told me that I could power motors up to 40v, though I would need to consider more surface area for ground so the motor driver burn up!

I think moving forward, I'll play around w Neil's board, and add things that allow me to control the motor and prototype the mechanics of the conveyor belt! Super excited to move forward - this somehow feels like a huge breakthrough, though its a small accomplishment to make a dc motor move w a board i didn't even design...

10 / 23 / 19 :  motor print test

10 / 24 / 19 :  testing the test

10 / 25 / 19 :  testing the correct wheel

10 / 26 / 19 :  julio's belt close up

10 / 28 / 19 :  aluminum or acrylic + color

10 / 29 / 19 :  scaled down version

11 / 01 / 19 :  teflon, joints, and funding

Today I met with Diego to discuss my final project, which was super exciting and helpful.

redlining the current section / notes :

Here are a couple things we discussed - some reference the drawn over section above :

+ teflon strip under lip :

although, I've added a height / adjust screw in order to not rely on perfect tolerances / and adjust the wheel as needed to put harder or lighter on the plates above, the plates might "jump" when they arrive at the wheel due to whatever slight change in height the wheel might be at. Diego suggested adding something that would push down on the plates as well - but doing so in consideration of friction. The lip Paloma suggested a couple of weeks ago was meant to hold plates in but not push down on them. So Diego suggested considering adding a strip of teflon below the lip that would control the plate from any kind of jumping.

+ plate joints :

in order to further alleviate jumping, plates might be "connected" - anthony had also suggested this.. I expressed to Diego that I want the plates to read as "clean" as possible (A) and that I wanted to keep away from having to mill plates - as there might be a version in which there are MANY plates. We discussed there might be a nice way to incorporate a simple 2D joint that could keep plates aligned (B), but that I could also push these joints to the edges (C), so they would be hidden by the lip of the belt.

+ layered fabrication :

the fabrication of this conveyor belt is something i have thought about but have yet to address - as mentioned previously, i want this to be aluminum but it might be smartest to jus make the outer parts aluminum. Diego and I discussed possible CNC most of the belt from lightweight MDF or plywood. He strongly suggested combining parts as more parts required more tolerances to consider. In making the 4' belt, although I could make it out of one piece (consider cutting a 4' circle on the onsrud, i am going to force myself to make it out of two halves or four quarters in order to learn and be able to apply this same logic to the ~8' conveyor belt. We agreed I should not depend on glue, but possible mill some kind of registration points and maybe 3d print pieces that would make sure all layers are aligning correctly.

+ prototype :

diego strongly suggested to begin prototyping as soon as possible. Today, I will be receiving my updates to the 3d printed motor from last week (there was a huge back up on the queue so even thought I submitted my parts last saturday, I am receiving them a week later). I think once this is more or less figured out, I will start water jetting / cnc maybe a quarter of the 4' conveyor belt - as there are sure to be erros but I also need a big enough section of it to test appropriately.

+ funding :

I expressed to Diego that this conveyor belt idea is hopefully going to at some point manifest in a larger ~8' conveyor belt. I have an idea about creating a lecture series around the belt, and diego strongly suggested looking at funding opportunities. He thought it would be great to make a successful 4' conveyor belt and use it to apply for funding to make the ~8' one.

These are the funding opportunities he pointed me to :

+ CAMIT Grant by Council for the Arts at MIT
   $500

+ Bill Mitchell ++ Fund by MIT SA+P
   up to $2,000

+ Harold Horowitz Student Research Fund by MIT
   SA+P
   up to $4,000

+ Keller Gallery by MIT SA+P
   not a funding opportunity, but possibility of
   exhibiting the finished conveyor belt

It turns out, the Mitchell and Horowitz applications are due January 3 - so this might work out. Unfortunately, the CAMIT Grant for this semester has already passed - Diego suggested I might apply to this one for funding for my 4' belt. But I am anyways, excited about this, which I hadn't really considered before now.

11 / 03 / 19 :  motor assembly working

motor assembly prototype :

after a week of waiting for my corrected / updated 3d prints, i was able to re-test my motor I started working on in week 07.

I think I will resend to print one last time, but mostly to shorten things here and there. The pieces are a bit tight so the motor, for example, it difficult to insert, but once inserted is pretty secure. The rod at the opposite end doesn't want to enter the wheel completely bc of tolerances, but again, the tightness is sort of needed, so I don't want to make this thinner and wobblier.

screw detail needs to be added :

Ah - I will also have to add the screw heightener detail to the motor assembly since that wasn't included at this point.

!!! :

The L bracket that holds the motor could be a tiny bit tighter - if you look at the video below you can see the force of the motor moving sort of incrementally pushes it out of place more and more.

I am super excited that this is more or less figured out - I mean, at least at this stage. I'm sure I'll have issues once I start prototyping the conveyor belt, but this is v cooooooooool!

11 / 07 / 19 :  push pedal prototyping

push pedal assembly :

push pedal parts :

this week 09, we we asked to work on an input device. My final project is pretty simply in terms of electronics, but i knew i wanted a nice push pedal to control my conveyor belt. While my attempt to use step response via arduino ide was not successful, i was able to 3d print a pedal prototype.

printed push pedal :

Ideally, I would want to make this a nice push pedal - I just wanted to test whether it would work, and as is it is tight:

too tight of a fit :

Anyways, I wanted to make this out of aluminum, but I don't want to get ahead of myself. Maybe I should just make a very nice spring loaded button - or a very tiny finger pedal.

11 / 14 / 19 :  controlling a motor with a button

WEEK 09 was pretty rough. So was WEEK 10. However, by the end of WEEK 10, into the next week, I was able to successfully program a fabduino to control a motor with a button :

button controlling motor via fabduino + 2 breakout boards :

This is basically the extent of electronics I'll be using. I do want to make a very beautiful interface, which I hope I can start to focus on in conjunction with beginning to prototype.

I will probably add the ability to turn the motor in the opposite direction, and for WEEK 11 I will be using wifi. It seemed like the most possibly relevant to my project - so idk would be kind of cool to be able to move the belt through a website if this conveyor belt maybe became audience interactive (I have this fantasy of the conveyor belt being used to have lectures, similar to how Julio Torres delivered his standup - so maybe audience members could control the belt to point to specific objects! COOL).

11 / 15 / 19 :  controlling a motor with two buttons

Today I milled two more boards : another button board and another motor board. I left my second dc motor at home, but I was able to test controlling the motor with two boards.

I think I'm a little careless when it comes to hooking up these boards. I couldn't figure out why my new button wasnt working and I resoldered and ever considered milling a new board. Turns out I had plugged something incorrectly... Sigh.

Anyways, when I did finally connect everything correctly, it was very exciting (though I know it is incredibly simple) to control my motorized wheel with two buttons :

forwards / backwards buttons :

I will say that things are starting to get a lil crazy.. So many wires everywhere. I do kind of want to make a nice plate or circuit board holder to nicely orgnize everything. Maybe next week I will finally purchase aluminum or just prototype the belt w plywood.. I need to do it! Idk why I keep holding of. I suppose I am just barely able to be on top of weekly assignments!...

Here is the code i used :

const int pwm = 3 ; const int in_1 = 8 ; const int in_2 = 9 ; const int buttonPin1 = 10; // change this const int buttonPin2 = 5 ; int buttonState1 = 1; int buttonState2 = 1; void setup() { pinMode(pwm,OUTPUT) ; //we have to set PWM pin as output pinMode(in_1,OUTPUT) ; //Logic pins are also set as output pinMode(in_2,OUTPUT) ; pinMode(buttonPin1, INPUT); pinMode(buttonPin2, INPUT); } void loop() { buttonState1 = digitalRead(buttonPin1); buttonState2 = digitalRead(buttonPin2); if (buttonState1 == LOW){ digitalWrite(in_1,HIGH) ; digitalWrite(in_2,LOW) ; analogWrite(pwm,255) ; } else if (buttonState2 == LOW){ digitalWrite(in_1,LOW) ; digitalWrite(in_2,HIGH) ; } else{ digitalWrite(in_1,HIGH) ; digitalWrite(in_2,HIGH) ; } }

11 / 17 / 19 :  prepping prototype files

Today, I finally sat down and started trying to figure out how to make a prototype of my conveyor belt.

One thing I had discussed with Diego a couple of weeks ago was that assembly would be important to consider. How does everything line up? Glue? Alignment bolts? Etc...

So, for this prototype, I will making a quarter of the conveyor belt by milling .5" plywood. I will be making the quarter out of two eights of the conveyor belt, but rather than make 2 separate eights, I'll layer the pieces to "stich" them together.

I could admitedly make this conveyor belt complete (4'x4' plywood is not a problem), but having the larger conveyor belt in mind, I do want to make I could in the future make a larger conveyor belt.

This probably doesn't make too much sense, but I hope pictures of the assembly will clarify it.

plywood assembly :

motors and bearings added :

aluminum plates added :

threaded rods added :

Like I said, i am not sure whether this will work, but I suppose that's the point. I did, however, purchase some parts from McMaster that I think I will be able to re-use regardless of whether this prototype is successful...

5" long alum threads :

aluminum cap nuts - for the top of the belt :

aluminum hex nuts - for the bottom of the belt :

4' x 4' plywood cut sheet :

Before I set off to mill this, I'm going to measure the stock - soemthing tells me it wont be .5" exactly, which will mess with my tolerances.

plates / new groove added :

in addition to milling the plywood, I'll be water jetting a quarter of the plates. I'll be testing the kerf to make sure the plates are cut correctly.

plates / new groove added :

As you can see, I've added a groove to interlock plates - Diego and I also discussed this during our check-in a ocuple of weeks ago. This groove wont be seen due to the "lip" on the top most later of the belt.

11 / 18 / 19 :  prototype redo

Today I visited the MIT Machine Shop as well as N51 to take a look at prototype materials.

Yesterday I started to set up files and discovered I would need :
+ one 4x4' sheet of .5" plywood
+ one 2x2' sheet of .25" aluminum
+ one 2x2' sheet of .125" aluminum

After visiting the shops, I found out aluminum is pricey, so I should probably hold off on it until I know what I cut will work. These are the prices :

+ one 4x4' sheet of .5" plywood $55
+ one 2x2' sheet of .25" aluminum $100
+ one 2x2' sheet of .125" aluminum $50

Because I would be holding off on aluminum, I decided I would use acrylic - which I had mentioned using. So, I took a trip to Altec Plastics in South Boston. There I bought :

+ 32" x 18" sheet of .25" clear acrylic $27.60
+ 32" x 18" sheet of .125" clear acrylic $15.00
+ 32" x 18" sheet of .25" red acrylic $33.50

After buying the material, I also measured it and of course it wasn't exactly the dimension it was sold as.

As a result, I am going to remodel the prototype with the actual dimensions and try to pack as much many cuts into the stock since it isn't cheap.