The original goal of our studio was to create a robotics kit for students at NuVu to use over the summer. Each group would work on a module for the kit I decided to try to create wireless connectivity for the robot. Originally I wanted to go for WiFi and Bluetooth connectivity I was thought I might create an app that allowed the interface to happen from an android smartphone and the Arduino controlling the robot. I ended up just using a Bluetooth terminal app I found that was already made.

For the first couple of days, I played with the Wifi module and looking at how the code works I realised it would not have been realistic for 3 weeks because of all the internet protocol I would have to learn so I decided to just do Bluetooth. I had no idea how tedious writing a communication protocol is coming in I didn’t think about how every single word needs to be parsed in the protocol. the way the protocol works is with tokens so a word corresponding to another module and the function that module is doing and the delimiter which  goes between the tokens to end the token and start a new token dependent on the previous token. Because of each project needed separate functions I had to write multiple tokens for every project this took multiple days to do partly because I was stuck waiting for other projects to finish their code.

The original goal of our studio was to create a robotics kit for students at NuVu to use over the summer. Each group would work on a module for the kit I decided to try to create wireless connectivity for the robot. Originally I wanted to go for WiFi and Bluetooth connectivity I was thought I might create an app that allowed the interface to happen from an android smartphone and the Arduino controlling the robot. I ended up just using a Bluetooth terminal app I found that was already made.

In this studio we were given the concept of creating a creative robotic module. Each group would complete a separate type of module eventually leading to the connection of all six modules. A base hole pattern was given and each module was built to connect to the base robot.

Our module was the creation of a functional robotic claw. This claw was designed to pick up any object the robot may encounter. It was also designed to fit on any type of robot with the same screw pattern, due to the back screw plate. We wanted to create a simple yet functional design that could easily be recreated.

We wanted to create a claw that had four fingers instead of just two as most robotic claws function. Many robotic claws had been designed before, but did not look like the type we wanted to create. Our type of claw featured two of nearly identical types of claw. We wanted the two claws to connect together by having one vertical and one horizontal. They would intersect at ninety degrees to insure straight fingers and a tight finger intersection. We made an extra the addition of sensory pads to ensure that the robot applied the correct pressure when picking up objects, which would not be inappropriately hard and destructive in its touch. We  thought four claws were superior for it allowed us to be able to pick up a variety of objects including more awkward shapes, such as pencils, rather than only cubes and spheres.

Our claw consists of two individual claws each built off of hexagons. Each claw features two hexagons with a gear system in the center. The gear system is created of two large gears which mesh together and lead to the fingers of the claw. Then a hole was cut for the placement of a servo in the hexagons. The servo head was removed leaving a small gear which meshes with the gear system found in the center of the two hexagons. When the servo was activated to turn, it would cause both gears to turn bringing the fingers closer together. The fingers are connected in two sections along with the incorporation of connectors that fasten to the main hexagon of the claw. This makes the rotation of the fingers more smooth. Both claws are connected together through a notch system. There are two notches found in the back of the small claw and front of the bigger claw. Each of the two claws feature a pressure sensor found on the fingers. Through arduino we were able to create a code that caused the claw to stop closing when the pressure sensor was activated. This was the servos will not strip, and objects will not get damaged. The arduino is attached to the larger claw found in the back.

Our project developed slowly with frequent iterations and many failures. The first iteration focused mainly on a box to hold a servo and some gears on top with arms attached, but there were quite a few problems (like forgetting that servos have cables…) and it took many tries in cardboard before there was one that was the least bit presentable. Meanwhile, Ryan focused on a more beautiful design which incorporated many useful features, like ways that the gears have to connect to the main body. After Grace had learned how quite a lot about the servo’s nature, and Ryan had learned quite a lot from analysing other works and from his own structure, we put our ideas together to make what we are presenting today. It went through one last iteration, which included adding a second set of claws and a mounting board to attach it into the main chassis and one for the arduino. After that, we added the sensory pads and found ourselves at a satisfactory place for Friday/today.

There were quite a few, from the design itself to how to hold the sensory pads on. One in particular was the cables. Towards the end, we found ourselves with an abundance of cables, but no where to put them. They looked hideous and we were fearful that they’d get caught in the fingers. Another was the fact that it had to be made easy to use for the younger generation. That meant if we went too complicated, it would be too difficult for them. It wasn’t too much of a constraint, but it did ring in our minds a lot. Then finally, we had some trouble with the sensory pads. If we had longer, we would’ve loved to edit the way they were attached. At this moment in time, they are fixed in place with electrical tape, but we would love to change that.

Our first iteration was a box that contained a servo with gears and fingers. It worked quite well and was able to rotate both fingers, but we had to play around with the location of the holes. This iteration was important because it was our first, and taught us about the limitations of a servo. In this iteration we had glued a gear directly on the top of the servo head. It worked well enough to turn both gears and activate the fingers, but the gears often slipped. We realized we had to find a new was around this problem.

After the completion of our first iteration, we began researching more designs that were inspiring. The iteration featured a design that was similar to our final design. It featured a single hexagon base plate with dual gears and a two part finger design. The fingers also featured a connector system which provided much more stability and accuracy of the arms. The connectors connected to the base hexagon. One of the biggest updates we had was the updated gears and servo. We removed the servo head which revealed a gear. We found the exact gear pattern and incorporated this into the gear system. In this update, the gears would not slip.

Our final iteration featured a much more advanced design compared to our previous projects. This design featured a highly modified hexagon base. In this new base, a section was cut out for the servo, and a top piece was placed over the gears and bottom hexagon. This way all of the mechanical and moving pieces were stored between the two hexagon bases. The next major update was the addition of the second claw. This claw would go behind the first claw and had bigger fingers to accommodate for the distance issue. The two claws were designed to connect through a notch system. This way the two claws connected at a ninety degree angle and all four fingers intersected perfectly. We also had to make modifications to fit the arduino and store it on the back claw.

Our project was a robotic claw module that was designed to pick up objects of varying sizes and shapes without damaging the object. Our module was one of five others from other students and we would ultimately join modules together. This would lead to the creation of one large robot. We wanted the claw to feature four fingers instead of the basic two. We also wanted it to have a design that was somewhat simple yet creative, so that it could easily be reproduced. To accomplish this feat, we created two separate claws and connected them together through a notch system. The fingers of the claw were powered by servos which provided enough power to grasp objects. On the ends of the fingers we used a rubber foam to make gripping objects easier for the claw. In addition, we incorporated two pressure sensors on the ends of the fingers to sense when the grip strength was excessive. To program the servos and the pressure sensors we used an arduino uno. We then were able to program the exact movements of the claw and at exactly what force to stop applying pressure. We had to create several iterations before our design was complete. We had many failures misunderstood areas of the claw before we fully grasped the concept of simplicity. Through many iterations, we eventually created a functional robotic claw that could grasp objects with ease.

My module is aimed to give the robot personality and build character.  On this module, there are  two small screens portraying eyes that display pictures and show different emotions.  When I started designing this module, I had intended for it to be one larger display on the back of the robot.  This display would function as a smartphone and have all kinds of different apps on it that you could use.  As I started looking through the displays that I would actually be able to buy, I noticed that these kinds of displays are very expensive, and would have to run on a raspberry pi, (which we did not want.)  Soon, I started seeing really nice small screens, and I decided to do my eye idea.  The screens I ended up getting are 1.8 inches, and come with SD cards.  With the help of our app, you are able change the eyes of the robot by uploading new pictures.  These pictures have to be a certain size and number of pixels, but if they are, it is a simple process to get them on to the displays. The main reason for creating these displays was to have a module that connected the human to the robot.  All of the other modules were purely mechanic and had very specific purposes, but they didn’t let the robot show emotion.  The display module also lets you personalize the robot which might make it more appealing.  There are several major components to my module, the most important being the displays.  Each display is mounted on three pieces of wood.  The first layer of wood had spots cut out for the chips and the header, as well as an open end for access for the SD card.  The second layer has a hole hut for the header, and the third layer is the same.  Each layer also has a space cut out in the middle to make room for the t-nuts.  When the displays are attached to the stack, 10 wires soldered to the header lead down to a small solder breadboard, which is connected to the arduino.  When you put a command into the app, the robot will say what you want it to.  I started watching and reading all of the adafruit tutorials on how to use my display and how to wire it up.  The wiring was pretty complicated, so it took me a while to do that.  When it was finished, I started looking at how to put the photos onto the display.  When I figured out how, I started testing everything out.  The photos that fit onto the display had to be a certain number of pixels and a certain size so as I started choosing photos. I had to resize them and set them to the correct number of pixels.  I played around with that for a long time, and then started thinking about what I was going to do if my screen couldn’t play animations.  I decided that I would just make a gif of an eye blinking and display that.  The eye had to be drawn, so that was the next task.  I used a wacom tablet to trace my eye.  This worked very well, and looked really good when I was finished.   My main challenge was the display wiring.  There was a lot of work for me to do, but mainly I had to solder a strip of ten wires to a header.  It was just busy/repetitive work.  A  couple times one wire git cut short or burned up, and I had to redo everything.  When the displays were finished, we had to place them on the robot.  In the design we have, the modules come together to a mailbox like shape.  At first I thought that I would mount the displays on top of the front boards, and then I decided that I would put them inside of the wood.  I cut out a bunch of different models that were supposed to fit onto the displays, but they all had a little something wrong with them.  All had holes in the wrong places, and two didn't have enough room for the chips.  The holes were also to big for the screws, (the screws stayed in but they didn’t bite into the wood.)  I decided to remove the holes altogether, and manually make the holes when screwing down the display.  With all these new dimensions, I recut everything to fit.  As I assembled the new pieces, I noticed that the metal t-nuts were close to touching the back of the display.  If they did touch, the display would have shorted out.  To fix this, I took a piece of tape and stuck it to the back of each t-nut.  

    The Power Supply Module is the basis of any robot and serves as a power foundation for all of its electronic components.

    The module elegantly houses its components in a case inspired by those used in servers and can be slid into larger robots or mounted directly onto smaller ones. The power supply outputs both 3.3V and 5V at 3A from a custom voltage regulator circuit. Higher drain components are powered by a high-drain 12V, 60A output drawing directly from the dual lithium polymer batteries.

How Does Your Project Work: Our project works by an air compressor that is controlled by an Arduino which fills a PVC tank up with about 80 psi worth of air. Then a valve on the other side of the tank is triggered by the same Arduino and quickly releases the compressed air into the tube. The Air then travels around the tube to the barrel where a marshmallow or other projectile is rammed into like a musket. The air packs behind the marshmallow and pushes it out causing it to shoot with no fire or loud bang.

Progression Of Project: We both worked separately on different things, but we spent the first couple days brainstorming designs and finding basic parts for us to use. Then we settled on our PVC pipes and the length of the barrel is 1 ¼” by 2 ft long and the tanks is 2” by 8” long. After that we bought an electric air compressor from amazon for $20. Once we decided a rough Idea on how everything was gonna work we started building the clamps to hold the barrel and tank together along with researching what valves to use to connect everything. While the clamps were being worked on and the valves were on their way to NUVU we made a box to keep the compressor in and make it easier to connect to the barrel. Since the Arduino only takes 5 volts, our valve took 24 and the compressor took 12 volts, we had to make a breadboard with relays to control both parts. Then after everything got assembled the back tube was a tad short. That meant we had to distribute the curve so the tube wouldn't kink. That's how we came up with the big wooden circle that is zip tied into place. Then the final steps after everything was on was to download the code into the Arduino and mount it onto our main shell.

Technical Design Flaws: Some of our flaws were the fact that we used a big tube on the back instead of just using brass valves. We also bought to weak of a valve and had to jump to one that took twice the amount of voltage. Another thing we could of did for more power is made the tank with 3” by 8” PVC instead of 2” by 8”. It would have given us more air to pack behind the projectile and would come out with more force.

Challenges, Modularizing the Air Compressor: It was very important to have a container to house the air compressor because it would be very bad if one of the compressor’s moving parts came in contact with anything. The air compressor we bought came in a plastic casing, but it was bulky and was made mostly out of empty space. It also contained a flashlight, which was an unnecessary waste of space. We wanted our gun to be as compact as possible, so we unscrewed the air compressor and took it apart, separating the actual air compressor part from the unnecessary parts. Next I measured the dimensions of the now bare-bones air compressor, which proved to be more difficult than I anticipated because most of it was made up of strange curves and gears, and it was not all on the same flat plane. Once I got down the basic outline measurements, I roughly recreated a more box-like version of the shape in Rhino. I made the dimensions slightly larger than those of the compressor so I could be sure that there would be no contact between the wood and the moving parts. Once I got the shape down, I made some supports that would attach to the floor and ceiling of the box and would serve to hold the compressor in place. When I first started making the walls of the box, I knew that they should be more like a frame rather than a solid wall because we needed room to run the wires through and we needed to be able to see what was going on inside. A frame with a large hole in the middle would also improve airflow to prevent it from overheating. At first, I thought that a narrow frame would not be sturdy enough, so I made beams crisscrossing across the hole in my first iteration. I also made every face of the box tabbed so they would fit together better. I laser-cut this iteration out of cardboard, but unfortunately the beams did not leave enough room for the wires. I also realized that these beams didn’t do a whole lot to make it more sturdy, so I decided to remove them altogether in my next iteration. A few more iterations and small changes later, I had a prototype made out of thin wood. The air compressor fit well, but the supports were not enough to keep it from being jostled out of place and the hole I made on the ceiling to show the pressure gauge was in the wrong place. I fixed both of these problems in the next iteration by moving the hole in Rhino and widening an unused screw hole in the compressor with a drill. I drilled a hole in the floor of the box and ran a screw through both holes, securing the compressor to the box. Then I integrated Hudson’s clamp design into one of the walls and secured it to rest of the gun, finalizing the box.