web analytics

— urbantick

Archive
March, 2009 Monthly archive

Some first graphs to analyze the tracking data in some other ways. It is now almost one month (I have not yet the data from all the participants for this time period) of tracking and I think this is gona be the first milestone in the project. At the moment I think that one month could make a pretty good unit as a base to start analyzing the data. A cycle of a four week patterns could provide enough data to paint a rough picture of the activities and range. As it is the first month I m only guessing here and will have to check this assumption as more data will be coming in over the next couple of weeks.

The graph visualization focus on the quantitative aspect of the data together with the time information over the location information. The idea is to look at the schedule information contained in the record. This is of interest as the project is interested to enhance knowledge on personal, spatial routines. The graphs are visualizing the amount of activity over a specific time period. The periods are one day – 24-hours, one week and one month. Using these units of general time frames helps to establish an appropriate framework for the data. Participants are all understood to use these time frames. More specific units could relate to religion, culture or specific responsibility or job. These will be respected on a more individual level of analysis. In the graphs the x-axis represents time were as the y-axis refers to amount of activity. This is measured by the number of log points the GPS device has stored for the time period in question. The graphs do not give information about time spent in one location they solely focus on travel time between destinations.
One month analyzed by day and participant. In total there is four peeks over four weekends. They generally do match, although one peak has slightly moved into week three. This was the UK midterm week, a holiday brake. Participants who have children or work in a school have spent more time traveling during the normal weekdays. Surprisingly the Sunday at the start of this mid term week is very low. All of the participants have recorded little activity. It must have been really bad weather and people stayed indoors. On the contrary, one Saturday pops out extensively. It turns out that one of the participants had an intensive outdoor sports day, during that he generated a large number of points.
UDrecords_month_090305.BWomW1MU957W.jpg
Image by UrbanTick for UrbanDiary – Graph UD first month

There is an activity accumulation on Saturdays. This shows up in particular in the week’s graph. Saturday has more than double the amount of points over other days of the week. Not only this one participant who is doing intense sports activity on Saturday, but all of the participants tend to have significantly more activity on Saturdays. Other than that the weekdays are fairly equal in terms of activity with tendency to a low point midweek.

UDrecords_week_090305.QqS7eVepqhth.jpg
Image by UrbanTick for UrbanDiary – Graph UD week

Compared to the regularity of the week, the 24 hours graph shows a number of peaks. The graph starts at midnight with an expected flat bit representing few activities. In the first hours of the day there is some activity but it reduces to virtually zero in the early hours of the morning. The day then starts with a first peak of the morning rush hour. Around seven participants start leaving the house, but it then really takes off from eight, peaking around nine and coming to a first low point around ten. From this low the second peak starts rising immediately. By looking closely at the participant involved in these first two peaks one can see that actually there is two groups, one generating the first “rush hour peak” and the second group mainly contribute to the second similar peak about one hour later. 
The second peak has a twin peak with a first high point around 10h00 and a second one just before lunch around 13h00. After lunch around two o’clock there is the low point of the day with the least activity during this 24-hour day apart from the early morning hours. 
After the lunch brake, there is a fat afternoon / evening peak. This is representing a number of weekend afternoon activities like the out door sport that was mentioned above. Included into this fat peak are a first evening rush hour high point between five and six and a smaller second peak around eight, probably pointing to the visit of the pub after work.

UDrecords_day_090305.A91tqDS1hy4g.jpg
Image by UrbanTick for UrbanDiary – Graph UD day

Colour correspond with the key on the map here
Generally this reassembles the expected daily routine pattern of a western city. Surprising is more the accuracy the pattern shows up, rather than any unexpected results. Although the sample is not representative this was not expected to find this regularity.

Read More

An update on the collected data from the Urban Diary project. It is already three weeks now and it is still going. The collected data is good, the main problems lie in the handling and processing. I tend to focus on the 24h time frame visualization, where the data is replayed all in one day, rather than visualizing it day by day over three week. It is much denser in this way and patterns show up more clear. On the other hand the danger is that one off activities have a very strong influence on the visualization. Each participant is represented by an individual colour.
For this weeks visualization I again used the Google Earth but without the satellite imagery. So it is visually simpler and there is more control regarding the colours. I am also using the 3D Virtual London model developed here at CASA to provide some context. Having this background moving into analyzing the connections between the activity pattern and the morphology is one step closer.
The 24h cycle I have also changed this week. I have noticed that altogether there are activities roughly between 06h00 and 02h00 in the morning. The default duration on Google Earth is obviously 00h00 to 24h00. The recorded animation now starts at 04h00 in the morning and continuous until 03h00 the next morning.

UDrecord_all_090227 from urbanTick on Vimeo.

A zoomed in version of the animation visualizes the area around UCL. It is again replayed within a 24h time frame and representing the different participants with different colours. The normal workday pattern starts showing up again, 09h00 to 17h00, outside this frame there is very little activity.

UDrecord_UCL_090227 from urbanTick on Vimeo.

The third zoom is looking at a neighborhood area where participants live. In this case the colours used in this visualization are not based on individuals but they represent weekdays and weekend days. The darker purple is the weekdays, where as the lighter pink is the weekend activities. The emerging pattern tends to be focused on the main transport axis for bus travel and tube stations as locations. On the other hand the weekend pattern shows activities within the neighborhood and local streets rather than the big streets. So weekday activity tends to be towards the south in two time frames, one in the morning and one in the evening. The weekend activity, in this case, then tends to be towards the north and through out the day.

Read More

Ant_Closeup.n7atadTiiyGZ.jpg
Image – close up take from mrx.no

Ants use pheromones to mark their trail and guide following ants. They mark the path as they go along ant leave tiny little messages. If the trail is successful and more and more ants follow up the guidance becomes more intense and denser, whereas other trails fade out.
Exactly this was visualized by Sean Dockray in his animation Ameising 1.

Ameising 2 from urbanTick on Vimeo.

ameising1-middle.b6lLo15n32Je.jpg
Image – Animation still by Sean Dockray

The ants movement was recorded in a 45 minutes shot and then retraced with software support each ant, frame by frame (would probably be quite simple nowadays with the new After Effects functions).
The emerging output might not be the collective memory, as it is called by the author, but some kind of selective evolution of spatial organization.
To read more about ants, the new ant bible has only recently been published: The Super-organism: The Beauty, Elegance, and Strangeness of Insect Societies, by B Holldobler , Edward O. Wilson, on Amazon for some £30.00.

Ants from Kristofer Hagbard on Vimeo.

Read More