Just finished up my 2-week prototype Emission Bricks! Will be summarizing my key findings after I interview my participants; we’ll see if my biases/assumptions held up.

Conceptual Metaphor Testing

One of the largest barriers to energy conservation is understanding the concepts and terminology around electricity. Trying to decipher the difference between a watt or kilowatt hour, or even what either of those terms mean can be daunting for anyone. Furthermore, simply telling someone he/she has just produced 200 grams of CO2 emissions leads to blank stares, wondering just how smug you really are.

In an attempt to help people understand the complexity of electricty use, many designers use visual metaphors to bridge the gap. One attempt is GE’s data visualization of home energy use, designed by Pentagram. Using filters of watts, kilowatt hours, dollars, or gallons of gas, visitors can begin to comprehend the power consumption for dozens of their household appliances. Iconography, animations, and immediate filter feedback is a step in the right direction for putting complex into layman’s terms.

In my next round of prototyping, I will be using this site as well as my own set of metaphors for electricity and CO2 emissions to discover what works best.

(Thanks to Chris Cannon for the link!)

After a week of user tracking and documenting, I’m at the halfway mark testing out my Emission Bricks prototype. I’ve got into the daily routine of logging all the user behaviors and updating their dashboards first thing in the morning. For the remote testing, I set up an opt-in Tumblr blog and Twitter feed so participants can follow their own and others’ statistics. The prototype is going really well, but am finding that keeping all the data up-to-date is terribly time instensive.

For the second week of testing, I’m trying out a few different approaches with individual users:

• Tash (User 1) will receive an SMS everytime she checks-in at a location on Foursquare, alerting her of the CO2 emissions of that behavior. I chose Foursquare as it is her most active social media account, even though it has one of the lowest totals of CO2 emissions.
• Cooper (User 3) will receive an SMS at the end of the day summarizing the total CO2 emissions of all his daily social media activity.
• Jess (User 5) will receive an SMS everytime she takes a photo with Instagram stating, “The Instagram photo you just took released 17 grams of CO2 into the air. That’s the same as driving 1/13th of a mile!” (I’m trying out comparisions to behaviors that we all can relate to).
• All participants must subscribe to the Twitter feed, updated once a day with running totals.

With Emission Bits, I’m tracking a running total output of social media behavior and subsequent CO2 emissions. In different approach (if given the time), I should test for limiting behaviors. Right now, there are no ceilings for how much one can use Twitter or Facebook. However, setting a fixed allowance that a participant may use for the week, say, “Only 10 tweets per week”, would hypothetically result in different behaviors.

Both approaches are very different, but valid, and can inform how I choose to build key interactions later in the design process. 

I’m designing and building my first prototype, with some inspiration and advice from Adrian Westaway of Vitamins. When designing for design research, Adrian suggested, “try to design a journey to take them (participants) on”. Yesterday, I created 6 Gmail accounts, 5 ifttt accounts, 50 ifttt tasks to send emails from 5 of the newly created Gmail accounts to 1 “master” Gmail account (This sentence could’ve been written in code).

All the repetition set up a prototype that tracks participants’ production and distribution of public digital content. Using the collected data, I plan to publicly display behaviors such as amounts of tweets, uploaded photos, and status updates with Legos. Yes, Legos, a physical embodiement of data and my childhood. With insight from my survey, I will also be equating each behavior with a CO2 emission, updating up the totals daily to the physical display as well as an online component.

With this prototype, I hope to test a few biases/assumptions:

1. The quantified feedback should positively impact participants’ production and distribution of online content.
2. The public display will create a “shaming” effect: first with the sheer amounts of conent being produced by each participant and secondly with the subsequent creation of CO2 emissions.
3. By observing each participants display, non-participants will have an increased awareness of their own online habits and CO2 emissions.
4. Incentive to conserve does not have to involve monetary motivation, and can be based solely on normative comparison to similar groups of people.

(The) societal shift to moving 1s and 0s instead of atoms and mass has the potential to significantly reduce our footprint on the planet and achieve a more sustainable model for housing the soon-to-be 7 billion neighbours we share it with. However, since the ‘cloud’ allows our digital consumption to be largely invisible, arriving magically with the tap of the ‘refresh’ button in our inboxes or onto our smartphones and tablets for immediate access, we may fail to recognise that the information we receive actually devours more and more electricity as our digital lives grow.

— Greenpeace

I came across an accessible infographic video on the carbon footprint of the Internet. According the video, the more you watch, the more carbon you will burn , with the video stating, “0.2 grams of C02 per second of video watched”. This is actually equal to Google’s estimates for a single search request. Despite the 24.8 grams of carbon that will be emitted watching this video (it’s 2:04 long), it does a great job explaining a concept few of us know about or realize as we surf along. Note to self: a system or service for online carbon offsets.

Above is a cybernetic model of TCP/IP protocol in the context of sending or receiving a 50 KB photo. The TCP/IP protocol functions as a comparator - a component of a closed-loop system that compares information coming from a sensor to the system goal. In the case of TCP/IP, the protocol checks if a data transmission (divided into packets) is complete and assembled in the right order; anything less than completion and the protocol can request for parts of that data to be transmitted again.

This exercise seeks to determine the what of my thesis, the content; it does not necessarily refer to the overall topic, but the actual category and detail of content so as to define the why, how, who for, who by, where, and when. This exercise is not a linear process where defining what first is necessary, rather to grasp exactly what is being studied, however granular.

On pursuing a thesis about the environmental effects of cloud-based computing, I need to better understand what I am measuring as well as the infrastructure (so I can determine where and when is the best point for intervention). The what in my case is data - little bits of 0’s and 1’s that live on your hard drive, and are subsequently stored and transmitted by remote server(s). The more data, the more energy consumed by the server.

Data is measured in bits and bytes (8-bits); you’ve most likely seen the data on your computer in megabytes (MB) or gigabytes (GB). When you send any type of data over the Internet such as an email, photo, or gchat message, your data is divided up into packets. On average, the size of these packets are 576 bytes or 4,608 bits, and consist of a header and trailer, with the data in between. You may or may not know that your computer has an IP (Internet Protocol) address - a unique numerical identifier for every device on a network. Even websites have IP addresses. The header of each data packet would contain information on the origin or sending IP address, destination or receiving IP address, and total size of the packet. The trailer of each data packet would contain information on how many packets there are and in what order to reassemble them back into the original data.

If I were to send a friend a 50 KB photo, the photo would be broken up into approximately 87 IPv4 (Internet Protocol version 4) data packets and then sent out across the Internet. The TCP/IP protocol checks if any packets are missing, request packets from the sending computer, and notify the sender that the transmission is complete. This operation of error checking is called cyclic redundancy checking and used by networked devices when sending and receiving transmissions.

Direct transmission of data, which can be ineffiecient and take time (think of a landline phone call), is an obsolete method of transmission for the Internet. However, due to the non-linear nature of the IP protocol, a Google search request for example is not handled by one server, but by several, to give faster, more relevant results. There is acutally a carbon footprint estimated by Google for the average search request: about 0.2 grams of CO2. Along with the power a laptop consumes, Mike Berners-Lee estimates a Google search creates 0.7 grams of CO2. Multiply that by the 200 to 500 million search requests per day, and Google searching actually accounts for 1.3 million tons of CO2 emissions per year.

References:

Berners-Lee, Mike. How Bad Are Bananas? The Carbon Footprint of Everything. Vancouver: Greystone, 2011.

“Bit” definition, Wikipedia, accessed December 18, 2011. link

Ethan Zuckerman and Andrew McLaughlin, “Introduction to Internet Architecture and Institutions,” August, 2003, accessed December 18, 2011. link

Greg Ferro, “Average IP Packet Size,” Ethereal Mind, March 18, 2010, accessed December 18, 2011. link

Hugh Dubberly and Paul Pangaro, “Introduction to Cybernetics and the Design of Systems,” January 2010.

“Network packet” definition, Wikipedia, accessed December 18, 2011. link

Swanson, Joe. Interview by author. Written notes. Cambridge, MA.,  November 20, 2011.

Urs Hölzle, “Powering a Google Search,” Google Blog, January 1, 2009, accessed December 3, 2011. link

The (carbon) footprint of the world’s data centers is currently the same as one-seventh of the U.K.’s footprint, or a quarter of a percent of the global total.

— Mike Berners-Lee

Going Back to the Future, or to July 2011

An Alternate 1985

In the movie Back to the Future Part II, the main character Marty McFly commits the ultimate snafu by leaving a sports almanac in plain sight of an aged version of his arch nemesis, Biff, in the year 2015. Old Biff then hijacks the time-traveling Delorean to travel back to 1955 to give his younger self the sports almanac from the future. Over the next 30 years, Biff uses it to amass a vast sum of money from gambling on sports, always knowing the winner. When Marty arrives back to 1985, he discovers an “alternate 1985” where Biff is his step-dad, mayor of his hometown Hill Valley, and owns just about everything. Way to go, Marty.

By comparison, Fiona Raby and Anthony Dunne from the Royal College of Art camp put forward the idea of “alternative nows”, offering visions of “how things could be right now if we had different values”. (Moggridge, 2006) Excluding Biff’s iron fist, their work remains in the noir, suggesting, for example, a reality where children grow meat to power their television. Notwithstanding Guy Montag knocking on your door right now, I’d like to imagine a current state where the Knowledge Navigator actually caught on and gestural interfaces – rather than a mouse – were our means of interacting with a computer.

Coupled with a growing momentum behind the internet with things, these themes formed an area of exploration of my thesis for about 4 months. The notion of creating new forms of internet-embodied objects as a graduate thesis is very appealing; rants about the need for more tangible interfaces along with explorations by firms such as Berg are evidence that interaction design can extend beyond the screen. Earlier sketches of my thesis included a built shelving unit that glowed when I got mentioned on Twitter (I’ve got 78 followers so not that often).

But as I focused more on the making physical objects, it became apparent that I needed to go beyond, as our chair Liz Danzico put it, “interesting explorations of an interaction design student”. I decided to shift my focus from investigations in academia to what I had outlined in July 2011, consumption.

Plunging into the Shonash Ravine

Staying on the Robert Zemeckis’ riff, Back to the Future Part III finds Marty stuck in 1885 with only one way to get out: get a locomotive to push his time-traveling Delorean up to 88 miles per hour, thus enabling time-travel (duh) to send him back to 1985. The kicker, apart from getting a locomotive to go that fast, was the Shonash Ravine cutting off extra miles train tracks, leaving little room for acceleration and error. Marty’s sidekick, the slapstick genius Doc Brown, calculated a point of no return whereby they must commit to reach 88 mph or plunge into the ravine. Spoiler alert: Marty makes it back to 1985.

Among many environmentalists, there is a consensus that a point of no return exists for Earth, where we have done so much damage to the environment that human beings can no longer inhabit the planet. Doc Brown knew the exact point of no return on the train tracks, but unfortunately, we cannot agree when or what that point of no return is for our planet. Bill McKibben, outspoken author of The End of Nature, offers a number of 350 parts per million of carbon dioxide in our atmosphere as the marker, and has founded a non-profit around the concept. As of October 2011, we are currently at 388 PPM.

So are we going to plunge into a metaphorical ravine? Yes and no. The ability for our air, land, and water to absorb pollution and then provide its bounty is debatable. Moreover, our behavior, particularly around consumption of natural resources, is so far removed from the extraction, production, distribution, and disposal processes that we have difficultly measuring our collective impact, let alone an individual one. Lester R. Brown of the Earth Policy Institute summarizes, “We are crossing natural thresholds that we cannot see and violating deadlines that we do not recognize.” (Brown, 2008)

Back to July 2011

Earlier this year, I drafted a thesis proposal that outlined my exploration for the summer. It stated:

“In great excess, we can consume digitally at near infinite levels which (I postulate) further removes us from the consequences of our actions. The removal of meaning from the actual object offers another opportunity for investigation on how we consume and ultimately experience these virtual forms.”

To put it plainly, the further removed from the consequences of our actions, the more we will engage in those actions. Pertaining to our digital consumption habits, there are little to no barriers to produce, save, share, and consume digital content. It’s even the M.O. of internet-based services to make sure our digital lifestyle is seamless and without barriers.

As we shift our content and communication channels to a digital format, we begin to loose sight of exactly how much data we amass. On a personal computer, it’s easy to notice how much hard drive space we’ve filled, but do you know how much data you have in your Gmail account? Facebook? Flickr? What about all of your online content collectively? One New York based startup, Dispatch, is looking to bring all your cloud-based content into once place; a benefit for those who need to manage their content, but not for those who want to know where their content is physically located. As John Thackera puts it, “These technologies are supposed to give us a clearer image-but by sanitizing the subject, they prevent us from knowing reality itself.” (Thackera, 2006)

This brings me to server farms or data centers or whatever they’re called. They make cloud-based computing possible and can be found in the form of a small stack in a work closet or come by the thousands, housed in a massive building in Oregon. What’s curious about these (we’ll call them data centers) err, data centers is they consume vast amounts of power. In 2010, global data centers “accounted for between 1.1% and 1.5% of total electricity use.” (Koomey, 2011) The industry recognizes the monetary, and environmental, costs involved with powering and maintaining such large facilities. Recent advances are making data centers more energy efficient, however, as more extreme “green” measures are taken in the location and design of new facilities, many others, old and new, still run on greenhouse gas-emitting fossil fuels.

Now for July 2011, thinking about the consequences of our consumption. How much power does it take to send an email? Consequently, how much carbon dioxide is produced when I do so? Thankfully, research has been conducted around this question, and Mike Berners-Lee, founder of Small World Consulting, even wrote a book on the topic entitled, “How Bad Are Bananas? The Carbon Footprint of Everything.” But do we keep building more data centers as our data cloud exponentially grows? What happens in 10, 20, 50 years? Are all my pictures and sent emails saved in a virtual shoebox forever? These questions and others help lay the groundwork for my thesis as I move forward with my research, and I can’t wait to get started. Again.

References:

Berners-Lee, Mike. How Bad Are Bananas? The Carbon Footprint of Everything. Vancouver: Greystone, 2011.

Brown, Lester R. Plan B 3.0: Mobilizing to Save Civilization. New York: W.W.Norton, 2008.

Jonathan G. Koomey, Ph.D., “Growth in Data Center Electricity Use 2005 to 2010,” Analytics Press, August 1, 2011.

Moggridge, Bill. Designing Interactions. Cambridge: MIT Press, 2006.

Thackera, John. In the Bubble. Cambridge: MIT Press, 2006.

Thesis question v.27.1

I’m getting closer to a solid thesis question that encapsulates the “why?” with the “so what?”. Here’s the latest iteration (oh yea, I pivoted):

“Does demonstrating the correlation of cloud-based computing with carbon dioxide emissions lead to a decrease in digital consumption?”