Some Initial Thoughts on QSAT Project

In ITP, Quantified Self

This week, Lela Prashad gave the class an overview of how the Earth is imaged by various satellite-based instruments. It was very helpful to get a foundation in the various satellites resources that are at our disposal and their respective spectral, temporal and spacial resolutions. We were also briefly introduced to MultiSpec, a program that we used to manipulate and process the satellite data. While it may be silly, it was news to me that the visuals rendered from the various satellites simply come from spectral data that is then mapped to RGB values. I don’t see myself directly utilizing satellites for my project, however, as Lela mentioned, manipulating the satellite visuals to tell a relevant story can provide much needed context.


I initially thought that I was going to do an extension of my WellTap Project from last semester and tackle quantifying people’s domestic waste footprint. However, after some thought, I think I’m going to take it in a different direction as I want to make sure I’m pushing myself to learn new concepts that are relevant to the class. Recently, I started talking to Matt from our class to figure out a way we could potentially collaborate. We both don’t have concrete ideas in mind but we do know that we want our project to focus on a physical device that collects data, as opposed to manipulating existing datasets. The starting point for our discussion was to look at different sensors that we could employ and extract potential applications that were relevant to the geo-spatial scale we’re talking about. Here’s a list of some of our thoughts so far.


  • Satellites
    • Seeing if we could detect reflections from water puddles, specifically at noon when the sun is directly overhead, to determine where there are significant drainage problems in the city.
    • We’re interested by using imagery, satellite or other, to identify he lights left on by businesses (i.e. midtown skyscrapers) at night. This is probably not very feasible.
  • Voice or Sound
    • We both like the idea of finding an instance where we can correlate sound or voices to map the overall sentiment of an area, negative or positive. This could manifest itself in a lot of different settings, so depending on the settings, we could provide various forms of visual feedback.
  • Infrared/Proximity
    • We were talking about looking at particular areas where it would be interesting to measure how long a person stood stationary in place in front of billboards or poster to measure effectiveness (this would need to be considerably more fleshed out as there are a lot of challenges and possibilities for unreliable data measurements).
  • Radar or Echolocation
    • Probably the most complete idea so far is to set up sensors to measure car speeds on various avenues and map that to car/pedestrian accidents in that area. Until the recent PSAs, it was very unclear to New Yorkers that there actually were speed limits. This could further communicate the need to drive at reasonable speeds.

We are not set on any of these and are going to continue to explore, but for now, the two ideas that I’d like to dissect a little more are the car speed management system and measuring sentiment using sound.


Submit a comment