After watching the first Blade Runner film, I became obsessed with the genre of sci-fi and the idea of a future full of robots, cyborgs, and flying cars. The film mainly tackles the concept of humanity, and whether or not a bio-engineered humanoid can claim to be truly human. I'm stretching the concept and applying it to an inanimate succulent; can a plant, given a consciousness, become more than just a plant?
Warner Bros. Pictures
The main functions would be the ability to move, sense light, sense moisture, and play audio. For movement, I used two DC motors driven by PWM using a motor driver. For light sensing, I used 4 photocells, one in each direction. The moisture sensor doubled as a switch, indicating when the plant pot is inserted (I'll get into that later).
I had to integrate the photocell readings with the outputs of the motors. The robot would turn until the highest reading is in the front sensor, and move forward until the average of all photocell readings meet a predetermined value. If the difference between all photocell readings are neglible, then the robot simply stops turning and gives up.
The speaker would play audio when the moisture levels dip below a certain threshhold, announce that it's quenched, or declare it's own death with a recording of the Tears in The Rain monologue from Blade Runner
I decided to have the base and the plant be separate. This way, the plant pot could be switched out, and I envision being able to detect the plant type with RFID tags and attributing a unique for each breed of plant. Additionally, by placing leads that ran up from the outside of the pot into the dirt, the pot would create contact with the moisture sensor only when inserted. I effectively created a switch that detected both moisture levels and when the pot is inserted and removed, as the reading would immediately drop to zero when removed.
This was the first time I've built a working electronic product, so there were definitely a lot of changes in my process that I had to adopt. I've learned that designing parts to house components requires precise measurements and knowledge of tolerances for the 3D printer. Electronics require careful reading and research of documentation, and extensive planning for the circuitry. Everything on the inside was either dangling or taped together, so a next step would be to build something more rugged and reliable.
I wanted to create robots that could interact on their own and generate new ideas.
Arguably, the most difficult aspect of this project was figuring out the hardware. I had no clue what I was doing. I just bought some parts, read up on the documentation, and broke up the functionality to test and debug bit by bit.
I started off with just learning how to use the IR remote library, and how to encode hexadecimal signals. I tested this by using a remote to turn an LED on and off.
I started to integrate components that I was actually going to use in my project. Using a solenoid and a ping pong ball glued on top as a makeshift head, I was able to simulate a bobbing movement by sending an IR signal.
By attaching both an IR light and receiver, each robot can effectively send and receive signals. The hardest part was figuring our a protocol for the two to communicate. In this stage, I just offset the sending and receiving by a few hundred ms, and had the robot wait for a signal once it finished sending its own.
First, I experimented with a soft-bodied design, where the movement was driven by mini solenoids in the robots. Most of the material consisted of an array of discarded objects, including socks, paper cups, scaffold wire, and ping pong balls.
I melted a hole in the ping pong balls and fitted an LEDs into the openings. In addition to a bobbing movement, the robots will now light up. I fitted all of the wires and the breadboard within the paper cup, cut two wholes for the IR emitter and receiver, and then tucked it all within a sock. I cut a circular opening in the sock and glued on a circular segment cut from a paper cup. This allowed the ping pong ball head to be snugly attached to the sock.
After choosing a form, I had to work out how my parts would sit within the 3D printed body. I measured each of components and started to model my parts in Fusion360.
This project taught me the difficulty that comes with working with hardware: debugging at times, would take a three-pronged approach that looked at my code, circuitry, and also the fidelity of the components themselves. I've found out that it isn't uncommon for sensors to be faulty out of the box! All in all, it took me countless of sleepless night to get everything just barely working.
In my future projects, I'd like to take it a step further and have everything soldered or learn how to design a PCB. Either the connections were shoddy, or there just wasn't enough power for the IR to operate over longer distances. As a demonstration, I set up the robots to send each other random musical notes to play. If I had more time, I would've liked to make it more evident that information was actually being sent via IR, and apply some sort of music theory, so that the notes being sent were not just random.
Dinner table conversations were lackluster in my family. Delving deeper into the problem and searching for opportunities to intervene, I recognized the importance of salt & pepper shakers in American culture. No matter the drama boiling under the dinner table, there's never harm to ask your family to "pass the salt."
We first had an exercise studying ergonomics and the hand. Using condoms full of plaster, we casted handles that we then sculpted. Observing the resulting forms, I designed a handle that was informed by the ambidextrous handle. These concepts would later be translated into the form of my shakers.
After creating the molds, I translated the forms onto paper. Finding geometry and fluidity in the forms, I then integrated them into the ergonomics of a ambidextrous handle.
Using a solid foam block, I marked the projected contours and cut the part out on the bandsaw. After sanding and filing, I was able to achieve a smooth and fluid shape.
With slipcasting, the timing during each phase is absolutely crucial, ensuring that I achieve the desired thickness and have it come out of the mold intact. It takes around 45 minutes to make one, and I poured around 15 of them only to salvage 6 in the end.
After cleaning the shakers up, I poke holes on the top and send them to the kiln. Two holes for salt, three for pepper. I tried some patterns on the top, like speech bubbles, and question marks.
After being fired in the kiln, the shakers shrink around 30% in volume. Accounting for the shrinkage, I modeled a plug with an expressive mouth and printed it out of flexible resin. Users would be able to twist the plugs to change the expressions on the shakers.
The final pieces were dipped in clear glaze, and immediately wiped clean to leave only a thin layer of residue. The resulting product, surprisingly, took on a beautiful speckled and matte finish.
Initially inspired to create something explicity be used as a productivity tracker, I decided to instead make a product driven by its user interaction and experience, leaving the functionality up to the user. The moodboard is an open ended device, framed to be used as a calendar and journal. Users encode information through the act of placing colored wooden balls within the slots. Recognizing the lack of tactile response in the age of digital devices, it invites users to interact through touch. There needs to be just enough instruction so that users would know what to do and what could be done, but not have creativity be limited.
There's something about balls that bring out the playful curiousity in people; they beg to be picked up, felt, rolled. Most children experience some form of ball as their first toy. I wanted to channel the universality of the ball and communicate functionality without the need of words or images.
I've always loved Rube Goldberg machines and marble runs. The first time I encountered one of these was in the Stanford Pediatric Hospital. Forgetting whatever ailment I was burdened by, I was transported into another world, built upon the zigging and zagging of colored balls speeding on metal rails. The excitement I felt as they turned a tight corner, sat patiently as a hammer wound up behind it, and came to an exhausting stop just to be sent back to the start, lifted by a tiny elevator.
Hoping to emulate the captivation that I felt watching the kinetic sculptures, I wanted to create a spectacle every time the user operated the device. I initially wanted to create an intricate network of winding paths behind a grid of boxes, which balls would roll down to reach the next unfilled slot. The design became overly complicated, and I had to decide where to best concentrate my efforts in order to make the most intuitive, immersive, and rewarding device.
This meant that I wouldn't have to deal with figuring out how to get the balls to their intended slots, and how to then reset the board. Users would be able to place the balls themselves, and run their hands along the board to then remove them. With this approach, I could capitalize on the tactility of the balls and the wooden material, and allow intuitive, physical feedback when operating the device.
I first designed the CAD in Rhino.
Ideally, I would have made 3 of the complex parts on the CNC machine, however, I ended up having to build the bottom shelf by hand, due to the restrictions of the CNC machine.
The bottom shelf was built by hand, and the middle hole-punched board was cut out on the laser cutter.
I bought the wooden balls and spray painted them different colors.
Xondas is a company on the UofM campus founded by a group of professors within the EE and Physics department. They wanted to utilize magnetic levitation in the product to highlight the fact that they could transfer a large amount of power across an empty space. Being the first time I've worked on a project with a group of engineers for the goal of commercialization, I was extremely humbled and nervous. My work had to be much more meticulous than the scrappy prototypes that I made for class.
After much debate and discussion, we came to the conclusion that, with their goal being to start a crowdfunding campaign, we'd want a slick product that would capture the target market of young tech-loving adults. I pitched many ideas that would utilize wireless power, including a living boardgame, floating lavalamp, and electronic building blocks. For the sake of simplicity, though, we decided to stick with a floating lamp. Although the idea itself wasn't entirely new, Xondas' technology allowed 10 watts to light an entire room, whereas in competitors only a dim LED could be powered. Specs like this, though, are hard to sell, and ultimately the product would have to stand out with its look and feel.
After experimenting with a levitating magnet kit, I determined the ideal volume, shape, and mass that something would need to keep afloat. It can't be too large, or it'll wobble and topple over. It can't be too heavy, or it won't levitate very high. It's also got to be symmetrical. We found that by tweaking some settings, we could have the base float on an angle, exaggerating the effect of suspension.
Aside from the overall form, another aspect I wanted to focus on was the user interface. It needs to communicate functionality both simply and elegantly. The user must be able to understand its function at a glance. I looked at knobs, sliders, and different dial mechanisms.
Working with CAD can allow me to better envision materials and the effect they have of the light. Since this product will be a lamp, the way light bounces off of the surfaces, diffuses, and subsurface scattering are all extremely important.
Using a storebought kit to test the levitation, I revised my model and figured out how to build something that actually floats. Working off of my CAD explorations, I modified them to be actual parts to contain the electrical components.
Once I established a working based model and geometry, I continued sketching to create a more streamlined and appealing form. Multiple iterations were made to figure out a proper lid in order to minimize the material in between the magnets, and to figure out the perfect internals to support the components.
Additional detailing was added via a capacitive interface for user interaction. I experimented with different patterns of metal wiring on the floating body.
Throughout the whole design process, I worked in parallel with the engineering team to determine dimensions for custom PCB's for my forms. The challenge was that I had to first come up with a hypothetical design to understand what shape the circuitboard would have to be, be realistic in my demands, and then redesign the actual shape once I receive the final dimensions.
With the access to amazing rendering software, I've largely neglected to hone craftsmanship when it comes to building presentable prototypes. With this project, I tried my best to pay attention to developing practically informative, yet beautiful prototypes. In this version I used XTC-3D, and epoxy coating, to smooth the print. I then sanded the surface and applied a combination of primer and matter spray-paint to create an ideal representation of the material.
This year-long project gave me opportunities to experiment and learn that I couldn't have otherwise received from the classroom. It was the first step to learning how to develop feasible, realiable parts, through rigourous testing and collaboration. As a result, I've become more intentional with my design decisions, and learned the value of rapid prototyping as I became more knowledgable in 3D printing and materials. Most importantly, I learned how to collaborate with other disciplines, allowing my own expertise be enhanced by the knowledge of others, and to better utilize my voice and visual tools to communicate, negotiate, and realize a product.
This project was the outcome of Winter 2019 Interdisciplinary Product Development class, a joint course between the graduate schools of Business, Information, and IOE, and undergraduate School of Art and Design.
During the day-long design charrette, we worked with kids to map out their day-to-day activities, habits, and lifestyles. We worked within topics of physical activity, mental health, media use, sleep, and nutrition.
We found that most pre-adolescents are beginning to have their own digital devices. Phone use before sleep is a shown to be a key cause for poor sleep quality, and parents find it extremely difficult consistently regulating phone usage especially at bed time. We thought that this would be a perfect point of intervention.
After a whole round of concept testing, research, and a design review, we settled on our phone box concept. It would be a box that helps children develop good sleep habits by restricting media use before bedtime. A timer could be set on the parent’s phone, and the child’s phone will be locked until they wake up. We were striving to create a means for parents to passively regulate their children's phone usage, so instead of having to force kids to give up their devices, they would learn to put their devices away under their own power.
The central feature that I wanted to design around was the storage and retrieval of the phone. It must be intuitive, provide instant feedback, and also be rewarding to use. A few ideas we looked at were casette tape players, and how the tapes would be slotted in, and the lid would give a really nice click when closed. The lid would then sit flush. I really liked this aspect, because when closed, the device would be out of sight, and out of mind. The box would appear to be unopenable, until the timer goes off and the lid magically pops open.
I decided that this casette mechanism would not work with different sized phones. I ideated around different spring and lock mechanisms. I settled on having the spring and lock both on the front side of the lid. This way, the lid would pop open when the time is up, and the phone could be take straight out.
For our design review, I wanted to just focus on building a prototype that would lock when closed, and pop open after delay when a button is pressed. The challenge was to create a compact mechanism that would detect the lid, lock it, and also automatically open. This was the design that I settled on: on the lid, there is a potrustion for a servo to latch on, when the lid is pressed down, it compresses a spring, and at the base of the spring is a pushbutton that would then be actuated. This tells the arduino when the lid is closed, and the servo would then turn to latch the lid closed. When the timer is up, the servo unlatches, and the spring decompresses to pop the lid back up.
After concept testing and a design review, we decided that to make this a more viable and attractive product, we'd need a few extra features. We decided to have a digital clockface so kids would be able to see the time without needing to take their phones out. This meant that I had to add a slanted face, so this changed the overall design quite a bit. A UV sanitizing light and wireless charging would make the product essentially a smart device hub. Additionally, we added an adjustable nightlight.
I wanted to make sure the product was both fun to look at and to use. I used a light-up arcade button as the central point of interaction. Both the lid and the nightlight would be controlled by the button, so there was some trickiness to figuring out the code in arduino. Additionally, I was grappling with the limited space, and had to modify various components by sawing down the plastic and stripping down parts. We wanted to showcase the customizeability of our device, so I made two different colored versions, and lined the insides with different fabrics.
Our group has decided, with the help of the Murphy Prize for $5000 given at the end of the class, that we'd continue and try to commercialize the product. We've started by renaming our product, and incorporating under Remediate LLC.
A desktop device that would allow users to track and log their emotions, productivity, and other aspects of their life. The metaphorical mirror provides users the means to reflect on their lives and identify patterns of behavior, in order to make meaningful adjustments.
A prevelant theme in my work has been reflection and productivity. A lot of it has been geared towards better understanding myself, my habits, and how to optimize my productivity. The challenge is in achieving these goals through quantitative data that can be easily obtained from the user. Is there anything different that I could do than our smart devices? My solution was to build a device around the core idea of being able to intuitively store and display data, and in turn capture patterns in the user's behavior.
The more documentation I read, I the more confused I got. I had ambitious plans for my project, thinking that it'd be a breeze transitioning from arduino, but the raspberry pi was a while different beast. I gave myself one task: figure out the LED matrix that would be the centerpiece of the user interface.
I designed a friendly smiley face to display on the matrix. The hope was that I could create a state machine that would display animated gifs as feedback. Initializing the rpi matrix was difficult enough; after frying an SD card and spending days debugging glitchy output, I realized that I also had a faulty microcontroller. I switched to a new rpi and everything worked. Except that the test script wasn't able to display GIFs, and eventually discovering that every script on github was outdated, I kind of just decided to temporarily give up on the display.
I drew inspiration from the interface of the Nest thermostat. The entire body of the device was the dial, allowing for the means of interaction to be intuitively connected to the interface itself. However a problem I ran into was using a dial while connecting wires throughout the device. A slip ring would be required to maintain a stationary structure, rotating outer ring, and continuous wires running through. Added on to the complications of mounting the device and designing an entire interface with just the dial, I decided to meet my deadline by transitioning to a desktop, multi button and dial driven setup.
I had to print the parts pretty last minute for a showcase. The back of the base was only half-printed, as the 3D printer ran out of filament. There were a bunch of adjustments that I needed to make, as the cables and arduino couldn't quite fit in.
So, I wasn't able to get everything working on time, but I'm hoping to pick up this project in the near future, and implement all the functionality using a RPi and SQL to manipulate the data.
I initially wanted to create playing blocks that would be used to educate users about how the shapes and textures of sound waves affect the quality of sound. It eventually evolved into a more open ended toy that seeks to immerse the user with different forms of feedback.
Follow my progress on my blog here. Warning casual and sometimes obscene language is used.
How might we mediate children's device usage before bedtime?
Can machines create their own ecosystem without the need of human interference?
If Androids Dream of Electric Sheep... do plants ponder the prospect of world domination?
How could I track and identify my own behavioral patterns?
Brought on as a design intern, I had to build a product that would showcase the company's wireless power technology.
Could music be incorporated into playing blocks to enrich their interaction?
How could one promote communication within families?
Rediscovering the value of an analog experience.
Getting acquainted with 35mm film.
Graphic design for Dance Mix
Design and manufacturing