Author Archive

Wife Watcher –  John Choi, Haylee Jung, Sabrina Osmany

We wanted to do something interesting after sensing the presence of tags from the RFID reader. We decided to use the web and send a text message every time a tag was detected.

We used Socket.io and Node.js to get the data onto the web and Twilio to send a text message.

We developed the ‘Wife Watching’ story as a humorous take on what we could do with the technology. The Wife Watcher sends you a text every time your wife leaves the building. Enjoy :

WIFE WATCHER from hayleejung on Vimeo.


Read More
Adarsh, Ran and Sabrina


We noticed that the existing  model was killing off users if they did not find a vehicle. Therefore things were pretty perfect (for a number of different dock, vehicle and user permutations). As unserved users simply vanished, the model circumvented one of the major problems that MoD systems set out to solve (enough vehicles for passengers.)

We tried to count up the number of unserved users in a new variable: unserved. If the unserved users > 0, we incremented by one, each time.
After some brief experimentation, (with trying to get the unserved variable to show an incremented value on the inspect rollout) we resorted to simply removing the Die function. This adds up users to the user queue
We are keeping the rate of inflowing users untouched, i.e. random.

Variables we are Measuring:
-Total user queue per 100 ticks. Ideally this value should be 0, which would indicate that all user demand is being met.
-Served users per 100 ticks.

We are trying to optimize the rate of served users as a percentage of total users (served and unserved) per 100 ticks.
We then started to play with variables. Increasing the number of stations does  affect the user queue per station but it does not  change the user queue as they simply get divided between stations.
We decided to keep the number of stations constant at 5. Frame rate was constant and docks were also maintained at 10 per station. We counted served and unserved users between tick 100 and tick 200 (Because the first 100 ticks is when the system is just starting up)


Vehicles 5, Stations 5
726 unserved, 172 served. 200 ticks.
20% served
Vehicles 10
162 unserved, 446 served. 200 ticks
72% served
Vehicles 15, Stations 5
101 unserved, 932 served  200 ticks
90% served
Vehicles 20, Stations 5
281 unserved, served 1471 served
83 % served
At this point we decided to go back to our best result which was with 15 vehicles. We now decided to increase the number of stations from 5 to 8.
Vehicles 15, Stations 8
568 unserved, served 2286
80% served
Keeping the number of stations constant and increasing the number of vehicles, we observed an exponential growth in rate of served users.
However, keeping the vehicles constant (!5) and increasing the number of stations, decreased efficiency slightly. We think this is due to the fact that vehicles have to run between a higher number of stations (more targets) thus decreasing their efficiency.
Read More

This scatter graph visualuzes data about city bike users’ birth years (x axis) in relation to trip duration (y axis)   for the month of September 2013. This is done using D3 and underscore.js library. The example I used can be found here. I downloaded a csv file from citybike’s system data archive. I edited the file and got rid of most of the data (I only needed 2 parameters, birth year and trip duration) which made for a lighter and quicker load on the browser.

Screen Shot 2014-09-17 at 12.11.20 AM


I was surprised to learn that people born in the 1900s use citybike. The graph shows that the older population tends to have much shorter trips than the younger population.

Users born between the 1970 and 1990 tend to use city bike the most, and they also tend to take much longer trips.

I have capped the trip duration at 50,000 minutes because loading more data makes the browser crash. Additionally, there is some limitation in viewing the frequency of the plots when they overlap.

Read More

link to blog : http://www.itpadventures.com/cloudcommuting

1. In ME++ Mitchell talks about “electronic nervous systems” of intelligent urban environments. Discuss an example of an intelligent urban system you are familiar with and discuss the elements of the feedback loop, how its form of governance works, and who are its stakeholders (goals, decision makers, evaluators, etc.).

Reading ME++ has greatly brought into question for me the contours of modern urban man and where they meet the urban environment. Mitchell’s metaphor “my enclosures are leaky” is thought provoking.

With so much of our lives being tied up in networked interaction, it is no longer simple to classify a boundary that makes up a city, (conversely, a boundary that makes up the self) In the age of the internet, we live in an interconnected space and time. In this new spaceless and timeless polis we can observe one phenomenon that defines our  interconnected dwelling : the production of large amounts of information.

I am going to take some license in describing a system,that is not strictly speaking an urban regulatory system, but nevertheless an intelligent system that has great impact on urban life: The Marketing and Advertising Industrial Complex

Both broadcast media and online ads leverage information  by tracking user behaviour, in an effort to produce targeted advertising. This is a sophisticated and enormous project of of gathering and analyzing data, sorting and profiling the different species of audiences, deciphering their tastes and appetites, designing precise content and scheduling optimal delivery to effect timely injections of information that will persuade the specified user type to make a purchase- or at the very least to simply leave the unsuspecting subject teased with the consumer impulse that has come to be so inextricably tied to urban life.

If we look at this interaction as a system (even if the particular entities behind the effort are individual competitors- the drive and impact is systemic) we can surely see an intelligent system at work here. This entire study of urban dwellers (with the deployment of advanced tools and automated marketing) and orchestration of purchase schemes can be understood as an intelligent urban system, especially once we begin to take into account Mitchell’s ideas about what constitutes urban space in the connected world.

Goals, decision makers, evaluators:

It would be too simplistic too assert that this system is malicious or evil. The goal of the Marketing and Advertising Industrial Complex is to create consumer spending. This spending then leads to increased income and this in turn leads to further spending. The stakeholders are indeed opportunistic capitalists, but in the age of entrepreneurship, it serves aspiring entrepreneurs well to have access to large amounts of information and data analytics tools that can intelligently identify potential buyers and direct the traffic of content to those that would respond to the content- i.e. fit buyers to ads. Directing content traffic reduces mismatched or unwanted information (a.k.a spam) which helps everyone, so one could argue that all users are stakeholders in so far as they benefit from receiving personalized instead of irrelevant content.

The decision making here is done largely by the algorithmic mechanisms in place, the evaluation metric is the sales revenue.

Perhaps a more critical and overarching discussion about whether data driven content delivery simply predicts consumer patterns and choices, or actually affects them, may be worth having here. The question of the day then would be what kind of affordances does the Marketing and Advertising industrial Complex produce. Does being in this automated ecosystem of regulated content streamline the user experience, or does it, more insidiously, affect human agency?

2. Suppose that in few years we have driverless automobiles that can be used in shared schemes without the need for redistribution. Is this a plausible solution/future for you?

The proposal for the driverless car as a “consumer electronics device,” one which we can think of more as “networked computers on wheels”  is a very exciting potential future for the automobile.  Keeping aside all of the technical challenges that the project will face, the cost of the hardware, etc, the bigger challenge it seems, is the initial roadblock of the transformational period, as described in MIT’s report on driverless automobiles.

The adoption of driverless automobiles will require a cohesive plan that has all stakeholders on board. It is their ‘collective will’ that is required to surmount the initial hurdle. Negotiating a collective will between stakeholders (automobile companies, transportation operators, electric and energy companies, real estate developers and urban planners, information technology companies, and policy makers) with varying or conflicting priorities and complicated relationships is a hard challenge.

In addition, the nature of this project requires that it jumpstart quickly in order for it to be a usable system. Its efficiency lies in providing mobile convenience, flexibility and autonomy. This requires that there are many charging stations,  widely spread out at convenient locations,  i.e. that the full system infrastructure is in place. In order for that to be a reality, there need to be more driverless automobiles on the road, creating a larger network. The MIT report speculates that there needs to be a snowball effect in order to get the system running in a usable way.

This means that on the one hand there must be agreed upon standards for these automobiles made by different manufacturers, ensuring interoperability on both the hardware and the telecommunications fronts- this is extremely crucial as having closed proprietary practices will be quite detrimental to the initial snowball effect required for the transformational phase. Furthermore an initial large scale investment injection would help the driverless automobile grow its network at a quick enough rate for it become usable as a system ( with charging stations growth comes the convenience of flexible personal mobility which is one of the strongest premises  of this project ) Initial investments are needed and are the expected curve for the project to start up.

I think that if the growing pains of this project are addressed, and a negotiation between stakeholders is met, it can surely be the future of the automobile.

Read More