Walk All of Manhattan

From Personal Science Wiki
Jump to navigation Jump to search
Project Infobox
Self researcher(s) Alastair Tse
Related tool(s) pen and paper, Google
Related topic(s) Sports and fitness, Social life and social media, location tracking
Builds on project(s)
Show and Tell Talk Infobox
Featured image
Date 2013/03/27
Event name New York Meetup
Slides
UI icon information.png This content was automatically imported. See here how to improve it if any information is missing or out outdated.

Walk All of Manhattan is a Show & Tell talk by Alastair Tse that has been imported from the Quantified Self Show & Tell library.The talk was given on 2013/03/27 and is about Sports and fitness, and Social life and social media.

Description[edit | edit source]

A description of this project as introduced by Quantified Self follows:

Alastair Tse, an iOS developer at Google, recently moved to New York, and had a goal of walking every street in Manhattan. He tried a few different approaches to tracking his walks but nothing panned out, so he decided to make his own app that doesn’t use GPS or drain his phone battery. In the talk, Alastair shares his adventures in working towards the goal, and the interesting things he learned about himself from the experience.

Video and transcript[edit | edit source]

A transcript of this talk is below:

Alastair Tse - Walk all of Manhattan

So my name’s Alastair Tze and I’m actually also an iOS developer; I work at Google. This is a personal project of mine. I talked about it last year at some point, so it’s a little bit of a rehash, but then also some interesting stuff that’s come out of me doing this. So I moved to New York about a year and a half ago, so here’s my commute. I live or have lived in 27th Street and 6th Avenue and I would walk to Google to work, and this is the path I took. And one day I was chatting to my friends at lunch and they said, hey, do you think I could walk the whole of Manhattan and I could at least trek and go around all the grid patterns, and I could go from A to B and I could see which streets I could walk down. So sure enough I was trying to track where I was walking and so unlike all the guys who are really not into manual tracking, I started to writing it down manually like this, and I’m not sure if you can see but basically what’s happening I’m writing down which Avenues I’m going down, and which street corners I’m doing. Obviously, it really got too much and I couldn’t track what I was doing, and it’s not scalable. So I tried to find some other tools. I tried using Google Mind Maps, where you can draw your own maps, and unfortunately it’s sort of optimized for plotting points into a map or rendering (KMN? 01:44) file and not really optimized for me drawing these lines. So in fact, I drew more than 30 lines and it suddenly stopped working. I also looked at Googles latitude location history and unfortunately GPS is not always working. If I’m trying to figure out which streets I’m looking on, there is usually an error of say one or two blocks, so it wasn’t really a great way of doing and in fact you can see me jumping around here, this is me walking back home on a similar path and suddenly it has me somewhere else on another avenue. So I did what any good iOS developer would do, I wrote my own app to do this and you kind of imagine it as your own kind of notepad, but it’s optimized for putting points on to a map and plotting a route around those points. So what I’m doing here is in the middle screen is basically a summary of all the streets that I walked down, so accurate to which street or Avenue I’m on. The screenshot on the right is basically me plotting down these waypoints and the app is tracing a route down the street and trying to find what was the route I had taken, so if it wasn’t accurate I could add an extra waypoint and it would give me a good path. So this is the data that I have currently, I think I got from last night. So I’ve been doing this for about a year and a half now, and so even although the title of my talk is walk all over Manhattan, I haven’t quite done that yet. So it turns out Manhattan is huge and I didn’t realise that, I thought it was tiny but here is a really quick run through. So for instance, it turns out I never go up to Uptown or Upper East or Upper West, and in fact from this you can probably see that I’ve only been to Upper East three times and Upper West a little bit more. Very close to the Natural History Museum and then I usually walk over to the Shake Shack and have a burger. So I’ve done a lot of midtown because I live quite close to midtown and I walk a lot there, so you can see from this map the dot and the lines of how many times I’ve been down that route. So you can see there is very little gaps. Once you look at the data for myself, I can see that I always hang out in Lower East side, East Village and Soho, and so forth. But I never go to Tribeca and I don’t know what is down in the East over there. I don’t go to Downtown much and only at weekends for just taking pictures and stuff like that. But it’s a really cool sort of record for accidentally recording you know where I like to hang out and where I should actually explore. So I want to talk a little bit about what I have learned in the last year and a half, because when I was doing this it was a kind of a game for myself. I was new to the city and I wanted to see all of the streets, and I often wanted to go from A to B and I was always bored that I was walking down Broadway or always walking down a particular avenue, and I just wanted to explore a little bit more. But collecting this data made me aware of some of the things that I didn’t really realise when I started. But here are some insights. So the app is really manual, and your notice that I can show everybody after this talk, but it’s really manual and the question is always like why don’t you just track it automatically. So one, I talked about the accuracy problem and I only wanted to find data on which street corners I was going down. So for instance, I had never been down 29th Street until today and this was my first time, and I can actually use that data to figure that out and not just my memory. But it is also a system of manual entry, so I do have my current location, and what I have done is I have added bits and pieces on top of it to track and help me enter it. So for instance, in the background it will track some specific location changes. And what will happen is when I come to add the path I can call on those things to say, how accurate was that location you recorded about half an hour ago and do you know which street corner it was. So you can make the manual entry a lot easier by assisting it. I have Geofence reminders, so for instance, I can set up Geofences. So Geofences are basically a little lat/long, plus a circle, and basically if you go in and out of the circle you get notification. So I have made my app remind me every time to go to work and it will say, which path did you take to work and do you want to enter it now. Or, if I have been stationary for a long time, a long time being five minutes it will stop guessing that maybe you have sat down and maybe you are at a restaurant and why don’t you add your route. So the way the path is plotted is using Googles directions API, and what it does. It’s just a JavaScript of some web API and you can fire off a bunch of lat/longs to it and it will come back with what the best route is that it things you’ve taken. The interesting thing is that it actually gives you distances and which actual streets you are walking on. So I’ve been able to actually find out how far I’ve been walking every month, and finally the only other thing is that the data is only kind of interesting to myself. I showed other people this data and I get really excited about it, but other people are like oh well yeah, you know, there’s this crazy guy writing down everywhere where he has walked. So here is some stuff that I’ve looked at in my data, so I can find out distance travelled and I can find out which actual streets I’ve been on. So for instance, which is my favorite Avenue that I’ve been on and I can find that out. Also, time of day matters so for instance, sometimes I will go down a particular route and then find out it’s all dark, it’s 12 o’clock at night and I don’t see anything, and even although I’ve walked down this route I actually should come back to this street and there might be something more interesting during the day. Streets change, for instance, now I am biased about going down some new streets, but sometimes old streets are interesting and I haven’t seen a lot of it before, so it should really tell me which street I’ve been to a year ago and streets which I have been to 2 weeks ago. So here is one graft that I have quickly pulled out. The first blue bar is when I first moved to New York – and this is all in meters, so I guess this is around 70 km I walked in October, and you can see in November I started getting a lot more comfortable and I walked around 140km. Then during Christmas, there was a huge lull and I guess it was cold, and I may have been overseas. Then as I got back and spring came around and I walked a lot. You can see at the very end I started dropping off a little bit, and what happened was my girlfriend injured herself, so we couldn’t go out much at all. So this data is as I kind of so said it’s interesting to me, but also it made me motivated, like March is coming up and last year I walked 120 km, I didn’t really try to do that again this March.

So on my website. There is a really short video and I don’t know if you can see, but basically this is a cumulative aggregate of all the paths I’ve been on months. So you can see that towards the end I started not filling in all the gaps and also I wasn’t walking much. But also I did cover a lot. So this is all on my website, and that is the website that I have all of my data and you can poke around and have a look at it.

About the presenter[edit | edit source]

Alastair Tse gave this talk.

Linked content on this wiki

(The content in the table below is automatically created. See Template:Project Queries for details. If newly linked pages do not appear here, click on "More" and "Refresh".)

Project that build on this project  
We talked about this project in the following meetings