— ITPG-GT 2621 001 CLOUDCOMMUTING

Archive
Author Archive

I wrote a program which checks in Foosball teams and logs a game to a database using RFID, an Arduino, and a Python Script. My script only saves to a local database, and doesn’t upload automatically to a cloud-based server. I mainly used this assignment to practice reading serial information in Python, and to try and design a user experience with minimal interaction. In other words, I didn’t want the user to use anything other than his ID card to input all the necessary information for the game.

Read More

After much analysis and time using Vensim, NetLogo and Spreadsheets, I came up with a few numbers and rates. I don’t think any of these numbers are currently very meaningful, but they may be useful in the future analysis of the Citibike system. Here are the rates I found from the spreadsheets, where I simply compared a station snapshot from the JSON with the daily ridership data and inferred some numbers:

(from the last couple months)

–Avg trips/day: 34476
–Avg miles/day: 60712
–Total members: 123926
–Avg Sign-ups/day: 116
–Avg 24-hr passes bought/day: 1393
–Avg 7-day passes bought/day: 113
–Avg miles/trip: 1.76
–total docks: 11452
–Avg daily trips/bike: 3.01

The Vensim system diagram of causal analysis I was able to make was meaningless without the ability to run a simulation complete with rates and variable stocks and flows. I attempted to run a simulation, but had trouble getting rates with proper units to agree and play nicely. That will be a work-in-progress. My thoughts for further analysis using this method is to break down the boundaries into microscopic but data-sound chunks to ensure that individual pieces work correctly before inserting them into the overall system diagram. Here is a picture of the last draft I did before moving over to NetLogo:

Vensim

Finally, I moved into Netlogo to try multi-agent analysis. I expanded upon the provided model by simply creating some global variables to monitor different aspects of the system. I incremented a “deathCount” variable each time a user “dies” from waiting. By doing this in the waiting phase instead of the “pickup” phase, it ensures that I get a count of only those users who showed up to the station and didn’t get picked up in the acceptible timely fashion. I ran hundreds of simulations with tweaked numbers and found for obvious reasons that the most important ratio to serve the people in a timely fashion was a high vehicles/stations ratio. Here is an image of one of my trials and the resulting plots:

1C

Overall, much more focused tweaking of parameters and better design of the models has to be employed to pull meaningful data from the citibike system. For example, to improve on the NetLogo model one might not kill all the users at a station when a vehicle picks up a queue, but rather have a one to one pickup to simulate bikes instead of cars.

Read More

 Intelligent Urban Systems 

I would like to examine the NYC Parks and Rec system as an example of an intelligent urban system. I am a member of the NYC Parks and Rec, which allows me access to any of the gyms/facilities in the city ($100/yr) and offers free fitness programs classes and activities. In my opinion, this is one of the smartest systems a city can offer its citizens—affordable access to health and fitness equipment, and also to FREE and educational recreational activities in the many greenspaces throughout NYC. In terms of the various feedback systems in place, revenue can be generated by permits, memberships, and of course is budgeted by the city (to the tune of ~$350 million). The parks also employ of 11,000 workers in the summertime. The system, which attempts to increase quality of life for all NYC residents accomplishes this by continually converting unused urban sprawl or barren concrete land into park spaces. Partnering with volunteer organizations allows for much of the labor to be converted into “recreation” for those who enjoy the maintenance of greenspaces or civic charity. 

The measurable need of the growing population is more park space to use. The response from NYC Parks is to continually expand and designate new park space. The evaluation really comes in the form of usership for park spaces and membership numbers. Also, to maintain these numbers NYC Parks continually develops programs to train the next generation of park stewards. The more users employed by and volunteering for the parks increases the future security of growth and sustainability. 

MoD Dreams

Ideally when I want to physically go somewhere, if I need to use more than just my legs for walking, I would type in (or think of) my destination and appear there instantaneously. That would be teleportation. The next best would be a very simple transaction to hire a personal vehicle to take me straight from wherever I am at that moment to my destination in a reasonable amount of time. Interestingly, I might choose to use a bicycle sharing program for this type of transit even if a personal vehicle similar to a Zipcar exists. I wouldn’t use a personal on-demand automobile until it existed in a system without accidents, congestion, traffic, and other humans driving. The bike share doesn’t currently serve an area for me that makes sense for my commute which is the simple reason I don’t use NYC’s CitiBike. I own my own bicycle and ride everywhere, avoiding the subways as much as possible. Currently, my transportation cost in NYC amounts to the maintenance I have to do on my bicycle, and the rare occasions when I am going far enough to necessitate taxis or subways. This current pricing model for myself is much better than the options offered to me by NYC. For example if I wanted to take a subway to school, It would cost me $2.50, and I would be there in about 30-35 minutes. If I ride my bike, I will be there in 30-35 minutes and it costs me nothing. I choose a subway plan which lets me pay-per-ride for this reason. There might actually be weeks where I don’t use the subway, which eliminates the need for an unlimited subway card. CitiBike’s pricing is very good compared to that of the subway, but the nearest station is still too far from my house. The main problem with the bike share is serving me better than my own personal bike serves me, in terms of locations. If we can figure out this problem, I will be very happy. One of the best parts about having a personal bicycle is that it can be locked almost anywhere outside of an establishment, in much less time than it would take to walk from the CitiBike station to the destination. The ability to treat a Citibike a little bit more like a personal bike would improve the system. 

Read More