Dietmar Offenhuber
MIT
In the 1990s, we thought that the future would be virtual. No more travel. The death of the cities was predicted. Previously, Frank Lloyd Wright thought that telecommunications would spell the end of cities.
But exactly the opposite has happened. Cities are our future. Urbanization is continuing and accelerating.
The SENSEable City Lab
Not architecture
Not a media lab
It's in an Urban Studies dept.
Kevin Lynch theorized that there are 5 elements needed to map urban environments:
paths
edges
districts
nodes
landmarks
Urban studies, research, and proliferation of data:
There is an idea that data is just out there and all we have to do is visualize it. But also important is who collects the data, where and when is it collected, how does it relate to reality in general. How do we generate and use data in this context?
Connections, Venice Biennale Project, 2006
They took a real-time telecommunications data set. The data visualization is shaped by privacy concerns; the data is aggregated and corsened to protect people's privacy.
Phone data can be used to find out things about locals vs tourists, cars vs pedestrians. For example, if we know where the pedestrians are, we can send the busses to them instead of making the pedestrians chase the busses.
(Is it wrong that I get tired of how data visualizations all look similar? It's the tools dictating the visualization. More questions of representation and how our tools and our culture--and our tools as part of our culture--dictate what we see in a normative, negative way.)
The New York Talk Exchange
How does NY relate to the rest of the world?
(Erik Hogan's--not sure of name--visualizations are a bit fresher-looking. Cool that this is shown right after I said I was getting bored. It's like the good karma fairy was listening!)
How does globalization unfold on the neighborhood level? In Brooklyn, you can see how immigrant families use their cell phones. They did an ethnographic study on how immigrants use their cell phones. There are many social implications to this work.
Trash Track, 2009
This is an example where generating the data itself is very hard and expensive.
In this case, there was no data set. Global supply chains are highly automated and tracked, but at the other end of the system, for trash removal, this is not the case. Citizens know little about what happens to their trash. Even the professionals in the waste management field have spotty knowledge about how the system works. The data sets break down at the intersections when trash moves from one company to another, one process to another. This allows for abuses, like the international trade of electronic waste.
They decided to follow individual items from pick up to end. Using rfid was not possible because there is no ifrastructure to read and follow them. They used active location sensors which could transmit their location through the phone network, using cell id and gps. They recruited 500 volunteers.
Volunteers each gave 20 objects to donate in Seattle.
The tagging process was a nasty business, how to keep the sensor attached to the objects? They had to protect the sensors with insulation foam.
Volunteers then could follow the objects in real time.
They followed the items for 6 months and finished with a map of 3k objects. The electronics and hazardous waste travels most. Sometimes the trajectories are erratic. We see that some electronic waste traveled across the country and back to get to the same end spot as another piece of waste that traveled directly. This shows where there is room for improved efficiency.
There is EPA data showing landfills and recycling centers. When mapped, it mirrors population density and rural populations. They mapped the sensor data with the landfill data.
Another aspect was to observe the volunteers. They followed the items on their own. People understand very well how to read a gps trace.
Copenhagen Wheel, 2009
There are more bikes than people in Copenhagen, and the city is aiming to replace vehicle traffic with bikes.
They wanted to use bikes as sensors, but they didn't want to attach a lot of sensors to bikes or riders. They wanted something compact. So, they designed a sensor wheel that could be attached to any bike. The back wheel has a motor to support your bike efforts and also senses air quality, noise, etc.
You can't have air quality and noise sensors at knee level, so they are rethinking this thing now.
At this time there are 50 bikes in production. You can access your own data collected by smartphone application or web. You can choose to share your data or not. In return for collecting this data, you get real-time feedback about your behavior, road conditions, traffic, etc.
(want)
Models of data collecting are very important to think about.
