More photos

Improb Circuitry

A Newly Finished Drive Axle

Entering Aggressive negotiations with SIMON-

SIMON's backside

SIMON at critique

SIMON about an hour before critique- lots and lots of pieces

Reflections on Failure

It's taken me a week to distance myself enough from Smartsurfaces to write this entry without a strong desire to rant and curse more than usual. Despite my perhaps sullen appearance during last Friday's critique, I want to emphasize that I'm not as pissed off as I appeared; I'm disappointed.

I want this post to be as unambiguous and truthful as possible. We Failed. There is no getting around it, and anyone who thinks otherwise is kidding themselves. John has been saying all semester "the product is learning" and this was thrown around on friday as justification that our team succeeded. It is not. I learned a great deal this semester but I dont think it was necessary to burn through three-thousand dollars to produce a non-functioning thing, then argue that because we did just that in the process of learning, we succeeded. Our failure was one of wasted opportunity to see the realization of our concept.

Why did we fail? Were we over-ambitious? Did we spend too much time investing in concept instead of execution? Did systemic incompatibilities between members doom us from the get-go? Perhaps and perhaps not; I don't know for certain, Mr Phillips' surveys might clarify this in the future. From my perspective in the final days of the course, our group suffered most from incompatible investment. For Damien and Z, this is their senior design project, for Marc it was a pass-fail course.

Differences in Investment were most apparent in the hours group members were contributing to the project. Some members simply went home early, claiming they did not work past a certain arbitrary hour. Other's stayed all night to complete the task they had set themselves. Personally, I begged for extensions twice so I could focus on SmartSurfaces, and write my Russian papers on the weekends.

Towards the end of the week, this was becoming a blatant source of friction, making cooperation between members more difficult. Between technical glitches and lack of sleep, our group was suffering significant setbacks, coupled with the persistence of ambivalence we seemed to be burning at both ends.

The day of critique was pretty awful. We had succeeded in created the structure, but none of the "smart" elements were there. It was a big blow to our group morale to see zero synergy between the elements that had taken so long to construct, and the electronic brains of it all. After spending all night fussing with the new margaret board, Damien and I were adamant about it working. We wrote a quick bit of code to turn on the LED's, trying to breath some bit of life into SIMON's shell. It was a futile attempt-

...

I had hoped that when this project was complete, we would all have something to brag about, "I made that". Unfortunately, this is not the case. I feel fairly accomplished in having tackled some major problems and designing solutions despite the fact that they were not visible or employed during critique. Having invested so much time into this course, and given the rather disappointing outcome, I think some unapologetic bragging about what I contributed and learned in the process of attempting the final project is in order.

The Motor System:

Yeah, this was a pain in the neck. However, The challange was really rewarding. Damien and I worked together to create initial designs and worked through the entire process together. We produced drawings, 3d models, several prototypes, technical drawings and finally a functioning drive system. I am especially proud of being told the week of crit that our drive system would never work by a mechanical engineer at toyota, and proving other wise.

It Works- up your's Toyota!

Margaret Board aka Ghetto Peggy Board

I think for being put together the night before, this was a pretty impressive circuit. Not to mention we used only what we could forage, while the peggy board cost over $200, and for our use both served the same function-

CNC Galore!

Perhaps it functions to underscore the complexity of what we were attempting to do, but I am pretty proud of our group for making use of just about every bit of manufacturing technology we had at our disposal. The laser cutters were used to produce numerous gear mechanisms and drive system prototypes. I designed parts for both the 3-axis mill and the circuit board printer at A&D, and with the help of Rachel and Damien probably logged about 20 to 30 hours of production and tooling time on the mill. Marc Operated the WaterJet and Router in the FabLab to cut out our steel gears and SIMON's polycarbonate skin.

...

What I'm taking away from this course

• Trust is a two way door. It is not enough for part of a group to simply trust the other and expect work to be shared or even complete. Trust has to be established by the behaviors of group members, not simply added to the mix to solve underlying problems. Reciprocation is not a given in group work: you can get burned-
• Being more sensitive to other's ideas fosters group dynamic. Working in smaller groups or one-on-one with Damien for example, helped me to appreciate alternative approaches. In large group's it feels like everyone is vying to be the clearest voice, or you sink to the background. By being less insensitive during our large group meetings things went smoother and the sense of combativeness faded.
• Breaking down communication barriers between disciplines by finding a shared language. Again, I think this was illustrated most clearly by working with Damien and our use of Illustrator to communicate our ideas. Using visual language to communicate was much more effective than 'talking it out'.
• Simultaneously, while some of my most pessimistic notions of group-work have been confirmed, I feel much better prepared for future group-work in knowing what behaviors to avoid and which to pursue.

What was brought to the table for Critique

Simon

Simon is a heliotropic, feel-good machine. He interacts with a user through motion and LED color mixing and enters a default solar-tracking mode when not in use. Face recognition technology, a simple joint system, and a flexible LED array make possible a wide range of movement and interaction while photovoltaic panels provide the power. By fusing emotional connections with the user, we hope to transform the concept of sustainability from a condition of sacrifice to that of pleasure.

Team Softcore
Rachel Boswell - School of Art and Designn
Eric Harman - School of Art and Design
Marc Maxey - Taubman College of Architecture and Urban Planning
Lindsey May - Taubman College of Architecture and Urban Planning
Damien Stonick - Material Science and Engineering
Zilin Wang - Material Science and Engineering

Why I didnt sleep last week

The week began with Damien and I spending countless hours custom cutting gears and welding them into place. The process was excrutiatingly slow but the results were well worth with. SIMON(our smartsurface) now had a fixed mechanical structure that provided rigidity to the hinged planes, it could now stand up without flopping over like a dead fish. Damien also learned how to use the mig welder.

Damien's Practice welds

Negotiating with SIMON

Perfection!

Rachel and Z were pretty excited when we brought SIMON to Design Lab 1 in the early morning hours. Afterwords, we went to McDonalds to get breakfast and celebrate- Did you know they start serving breakfast at 3 am? Us neither-

The Next major roadblock came with the Peggy LED board. Whole days were arranged around working to assemble this board, the LED's and ethernet cable required over 4000 solder connection! The first attempts at testing Peggy were pretty disappointing to say the least; only a quater of the LEDS turned on. Multiple hours were then spent on testing all of the solder connections to the board and Z noticed that the IC chips were in backwards! As a group, we decided that enough time and energy had already been invested into the peggy board to warrant overnighting new chips to ressurect our lighting system. The Chips arrived early thursday and we were dealt yet another blow; only half the LED's were now working, and only just. The LED's flickered and fizzled, there had to be a short somewhere-

With hours to go, our options were becoming slim. Luckily Z had picked up some adhesive backed copper foil I mentioned to him earlier. With only a piece of acrylic and come copper foil, we (Damien, Rachel, Z, and myself) decided to scrap Peggy and build a new circuit. The result shown below was to laydown strips of copper on the acrylic for a makeshift circuit. The circuit used 4 TIP-120 transistors to control the circuits ground connections at 4 points, one for every color of LED we were using. The transistors allowed us to use the PWM signal from the Arduino to perform color mixing, and effectively replace the Peggy board. The whole process was fairly quick, with Damien stripping the ethernet from the Peggy board, Rachel then testing each salvaged pixel, then handing it back to Damien for positioning while I soldered the wires in place.

R.I.P. former Peggy Board

Creating our alternative board, I think we dubbed it "Margaret"

Completed Circuit with the few LED connections: TESTED AND FUNCTIONING

Here's to hoping Margaret wouldnt be a treacherous black-hole of time and energy like her predecessor-

Drive system: Success at last!

There's always time for...

...a good rant.

In response to an article here, which discusses the completion of the Yas Hotel in Abu Dhabi, two things come to mind:

1. Whoa thats really beautiful and they did that with LED's! Even Cooler!

2. That's a lot of LED's, I bet they were expensive. What else does it do? (nothing-)... huh, kind of a waste.

This hotel is ridiculous and I think it really bastardizes "Green" technology. So I suppose if this were my hotel and I had this absurd facade that I wanted to illuminate at night, then yes LED's are better than incandescent bulbs. However placing this extraordinary display on a hotel that is literally built around the burning of hydrocarbons (It is surrounded by a formula racing track-) is pretty disgusting.

You know what would be really great? Instead of blowing millions of dollars on this silly facade just 'cause you can, then surrounding it in pavement (cement being one of the most energy-saturated materials available), consider installing low-flow faucets. Maybe using ultra efficient washing machines to launder linens, saving thousands of gallons of water, would be a worthwhile thought for a city that lies in an extremely arid climate-

As the article points out, this application of technology is yet another example of excess without reason. Worse, if this were to become a trend with a technology such as LED's, it risks establishing a high-end market niche for industry as well as cultural connotations of luxury. Transforming a technology that may one day replace incandescent and florescent bulbs as commonplace to a gaudy signifier of the ignorant 21st century bourgeoisie.

Doomed Drive System

After brainstorming with Prof. Marshall and Zack for a while, we developed a strategy for salvaging as many components of the drive system as possible to keep water-jetting time for recuts at a minimum. Examining the all the dimensions and looking over the aluminum frame led us to believe that an error was not made in the dimensions of the components, but the frame and piano hinges were the source of our gear meshing issues.

The Group members that fabricated the frame are in no way to be blamed. They did the best that they could with the tools that they had access too. Unfortunately, the method of construction led to irregularities, albeit small, that compounded to make the alignment of CNC gears impossible.

...

A New Plan

The alignment of the axle and small gears within the frame is accurate as far as I can tell, so we're re-cutting the large gears, and customizing (grinding) each one to be placed in the center of rotation. It's excrutiatingly slow, but it will will be the best way to verify that the gears are meshing correctly-

...

Pieces of the Old Plan

Aluminum Rod For Motor Coupling

Final coupling after lathe work and tapping for set-screws

Assembled Motor Mounts (CNC'ed Delrin, Aluminum Motor Coupling)

with the filed and cleaned gears and support plates

A single complete Drive Unit

Installed Drive Units

GUH-

I HATE GEARS!

I have double and triple checked the dimensions and placement of our components within the frame, and everything is looking good... except the gears are not meshing properly.

It's nearly 6 am, I've been trouble shooting since before 2, and have yet to identify the source of our problem-

Drive System: Delrin Motor Mounts & Axle Supports

Hopefully the last night of milling! Rachel and I worked to mill out the delrin support structures for the drive system starting at 7 and ending at 3 am; another 8 hours logged on the roland mill-

Jigs for holding stock material

Cutting

Motor Mount

2 Axle Components: Axle Support & Endstop

Super-thin layer to keep parts secure during machining

Drive System Bits and Pieces

Tomorrow we will be water-jetting the first components of the drive system! Below are all the gears and reinforcing brackets to protect the aluminum frame from self-inflicted damage-

37489... 3749... 3750!

Yes, thats right, We have 3750 solder connections to make... and this is just for the LED's. We still have the wiring for the drive system, solar panels, CMUCAM, and a lovely assortment of microcontrollers to go-

This whole situation lead me to ask, "Why isnt there a coffee pot in Design Lab 1"?

Perhaps the algae team wouldnt mind lending us a few of their IV bags? Then team four could be on a steady drip of caffeine!

Damien, Rachel, and myself stayed after our group meeting to do a bit more soldering and made some decent headway. We managed about 50 complete pixel circuits with ethernet wires attached.

A Complete Pixel Circuit (front)

A Complete Pixel Circuit (back)

Tonight's Results

...and lots left to do

Circuit Boards

Our group ordered a Peggyboard to be used for emotive lighting on the belly of our robot. The PeggyBoard is a pre fabricated matrix of 625 red, green, blue, and white LED's. However, the board is only about 12" x 12" and the belly area of our bot is easily 10 times that.

Our solution has been to create a series of small pixel boards that will contain a cluster of 4 LED's (red, green, blue, and white). Ethernet wire will then connect our pixel boards to the PeggyBoard. This will allow us to disperse the LED's evenly across our surface, without sacrificing control of every single LED.

A Sheet of 25 pixel circuits

Cutting Test Group of Pixel Circuits

A quick Sketch

A quick sketch on how our drive system will look

A few notes:

• As decided by the team, our drive system should have multiple points of traction along each axis of motion. This will require the construction of an axle along the length of the hinge, A coupling to attach the axle to the motor, and multiple gear meshes (most likely a pair of 2).
• Delrin or some other self-lubricating plastic will be ideal for the support structure. The coefficient of friction between stainless steel and Delrin would make more complex ball bearings needless. Maintenance in the long term would also be reduced
• Gears could be cut on the water-jet.
• Exposed or hidden Gears? A few weeks ago, this issue came up and the team decided the folllowing: That hidden gears would be ideal, but since an alternative could not be developed, Exposed would be acceptable. Unfortunately not all of our group recalls this conversation. As a result, drive system design has been put on hold until further notice. Some team members would like to ask outside mechanical engineers for a second opinion.
• I'm really sick of gears ...and welcome someone else to try if they are unhappy with the results.

The Joy of Prototyping (Part 2)

So we've got our actual motors and are proceeding with designs for the drive systems. Damien and I whipped together this quick prototype to show how we might actuate our robot. We scored the finalized dimensions of the frame onto the acrylic so it would be easier for our teammates to visualize how these parts fit together.

I'm very confident that our drive system will be more than adequate to actuate our robot. The motors are rated at over 6,000 oz/inch, which basically translates to us trying to hold the motor in one hand, and squeezing a pair of pliers in the other hand in an attempt to stall it. We were happily unsuccessful-

The Joy of Prototyping

First Attempt Drawing

This hinge is a prototype design for experimenting with gearing and torque. The hinge and motor assemblies will be responsible for actuation of all six planes of our "Creature". Once we can demonstrate a successful test case, We will create a total of Five drive systems that will be self contained and installed into the completed aluminum frame. This prototype is not designed for our motors, rather to test some basic gearing principals and then break it (intentional overload) in an attempt to expose possible design weaknesses.

As you can see, I didnt properly account for the piano hinge and there is a slight gap between the gears.

Back to the drawing board....

Success!

Designers vs. world

I'm sure the language barrier illustrated below is familiar to anyone attempting to communicate to someone outside their area of study.

If all else fails, its funny-

Creature Feature (The Thang)

Team 4 (un-named thing)

Graphics Generated by Rachel and Damien

Not Blogging is a good thing-

I find that the more pressure we come under as each week passes by, the less we're blogging. For me, it's simply too difficult to justify the time it might take for me to articulate myself and carefully explain our "master plan" for creating a functioning SmartSurface. So instead, I've been devoting more of my time working to physically manifest the surface we've spent the last several weeks dreaming about.

For the next 17 days (yes, 17 days is all that remains!), I will not be blogging. Instead I will switch to "documenting" mode, relaying photos of our manic process. After all, a picture is worth a thousand words and frankly I just don't have the time to write the thousands of words necessary.

Henceforward I default to photos, with the occasional caption-

Long Awaited Task 4 Recap

Team Bonding(Finally!)

Objectives

This exercise prompts you to propose, develop, deconstruct and implement an idea based on the principles of an interactive, heliotropic smartsurface. The smartsurface concept should constitute a set of functionalities that otherwise exist in 3-dimensional space, collapsed into an ostensibly 2-dimensional space, thereby gaining additional functionality and/or appeal.

This exercise is intended to get you thinking about the scalability of your ideas and feasibility of their implementation. The brainstorming you do should be more structured, with the objective of developing enough ideas and potential pathways to pursue in the following two weeks. You should produce something that is either a tangible prototype, or a very clear indication (with detailed drawings, diagrams, perhaps computer models) of what you will be building in the ensuing two weeks

You should restrict your work to the tools and approaches covered in class thus far. This exercise values creativity in the face of constraints.

Results

This was the most difficult two weeks of the semester thus far- Our group did not work well together and a great deal of our time was spent resolving "misunderstandings". Nonetheless, in the last 48 hours we pulled ourselves together, set our differences aside and managed to produce a functioning prototype to was quite successful. I was honestly surprised, early on I had doubts that we would be able to produce anything.

The fruits of our labor were made all the sweeter by our difficulties.

The result was the creation of a surface that circulates a fluid to heat or cool an internal environment. Arguably, this has already been done by solar water heaters. Our idea is novel in its execution: To create a changing and dynamic skin, perhaps as an architectural facade, that can react by circulating fluid though a hive-like arrangement of bubbles. One possible scenario is illustrated as:

Pre-Dawn: the facade is in its empty state with the matrix deflated

Daytime: Once the outdoor temperature is greater than that of the building the bubbles are inflated. Throughout the day, the fluid within the bubbles gather thermal energy. The Structure of the bubble-facade lends itself to the added benefit of increasing the surface area exposed to sunlight. In this way each individual bubble within the matrix acts as a mini-solar concentrator.

Nighttime: Once the outside temperature falls below that of hte building, the fluid is removed from the Bubble facade and moved into the building, acting as a "thermal battery", storing heat to be distributed as needed throughout the building.

Alternative: This system could be utilized in multiple ways, including inversely to store the cooler temperatures of the night for use as cooling throughout the day-

Self-Critique: The largest personal qualm I had throughout the entire two weeks was that i felt as though we were taking "the long way around". Meaning, I felt that there was a better solution to the problem we sought to answer. Better, like this blog, is a complete matter of perspective' mine being from the standpoint of simplicity, efficiency, as well as efficacy. I think that our group became fascinated with a mechanism then sought to apply it in a positive way. In this way, we doomed ourselves to creating something that addressed nothing, at least, directly speaking.

Me, assembling the servo-valve controls

Johanna, attaching the fabric to the bubble matrix

2 Week Heliotropic Smartsurface from John Marshall on Vimeo.

Chronology Is Overrated

Yes, I know, you're sitting on the edge of your seat for the re-cap of last weeks critique... Well, you'll be sitting for a little while longer-

In the meantime, try to figure out where I can get one of these:

Design: Beyond "Pretty"

Carl and Max have an ongoing debate about design: is it a final step of of creating something, in essence the frill or finishing touches that make it"pretty", or does the design grow out of the act of making? Is it process or end result?

Design is both...and neither; it is a way of thinking. Over the past week, I have been becoming more and more frustrated with individuals who even after witnessing "The Future of Design" conference have a static definition of design.

I realize my way of thinking and practice of design are my own and no other's, but I find it completely unacceptable to work with anyone who wants to force their perception of design onto me, and worse yet, onto a group.

As a designer, I do not feel it is my duty to pick up the schizophrenic pieces of mechanisms and circuit boards and attempt to put them into a cohesive package. Worse yet, then make the package valid conceptually post-construction.

...

Rather than validate our efforts, this backwards process completely invalidates them. Beyond the end result, this practice is having disastrous effects on group cohesion. Without developing a key concept first, we have lost the "mental glue" that binds us together towards a common purpose.

Group Bonding... sort of-

The major challenge for our group design has been the construction of an inflatable matrix though which fluid can circulate. The desired affect being a surface that behaves as a circulatory system to regulate the temperature of a building in response to external conditions. A single bubble shown below-

Several issues have arisen from this design choice (among which are my own reservations regarding efficiency), mainly construction of a prototype as well as theoretical considerations for mass production. A heat-sealing process seems to be the industry standard for inflatables, but we are really lacking any sort of specialized tools to reproduce this. Armed with a soldering iron and a few yards of vinyl, we attempted to melt the sheets together. Early single bubbles were initially successful, but later attempts at the construction of a matrix of bubbles became... problematic.

The pictures above show our method of using a laser-cut chipboard template to guide the soldering iron. The quality of our heat sealing was awful. We had numerous burn marks and leaks.

Plan B: Breanna had initially suggested using the same glue as pool repair kits. We discounted this suggestion in favor of heat sealing because we thought the glue would be too difficult to control when we wanted a very grid-like matrix of hexagons.

Now the question was: how to utilize an adhesive and apply it in a controlled manner? Also, what kind of adhesive? As a general resource for glueing, I suggest using thistothat. It's a pretty fantastic website that can point you in the proper direction for glueing just about anything. Our options were to use either contact cement, or 3M 80, similar to spray77.

Application: Again using our lasercut template, I used the spray adhesive to bond the vinyl together, placing straws between each hexagon to maintain airflow once inflated.

We're exhausting all other options before attempting to use the contact cement because its pretty toxic stuff and outgases heavily, and it could take 7 days to dry completely!

Testing of the Adhesive-based structures to see if they are air/water-tight will begin as soon as they are dry.

I also continued to work on a heat-sealed version as I was waiting into the early morning hours for the spray adhesive to dry. After a bit of... modification to our group soldering iron, I was able to get a completely functioning matrix. Later testing should show us which construction method will be more durable.