This morning we had a fantastic set of presentations and discussions about intelligence and embedded systems in the built environment – thanks to all of the speakers and to Douglas, Christoph, and Bo for organizing the first deep dive. In our hands-on lab we have now assembled our wearable sensors (thanks to excellent preparation and planning by Josiah, Yi, and their team), and this evening everyone got a chance to configure their Arduinos, pachube feeds, and look over version 0 of the basic software load for the wearable units. In our first radio deployment we saw about 50% success with radios joining the mesh network and updating pachube feeds, and we have some theories to investigate tonight as to what was happening (and how we can make that number 100%). Our core arduino/xbee team of Raj, Jordan, Bob Evans, Robb Drinkwater, Jeff Burke, and Dan Ashbrook, have been tireless at debugging strange wireless modem and gateway behaviors.
Meanwhile, as mentioned in the lab time we have begun to form some Hands-on Lab Teams to push the sandbox into full swing.
Sensors – Majd Sakr, Robb Drinkwater, and Hubert Ley are the core of a team that will provide code for
interpreting the microphone and accelerometer data. They have already written the basic
functions for these and will be looking at thresholds for sensitivity, etc. As part of this, we are going
to have the put a microphone in each space (on one of the mesh routers) and start collecting
data as early as tonight so that they can examine it and see if we can set thresholds to (for instance)
differentiate between a group discussion and a presentation in the classroom area.
Interactions- Paul Domagala and Bo Wrodda are recruiting a team to look at how we use the button and
LEDs, for instance implementing a voting mechanism with the button, actuating an A/C power outlet
(we have one Digi relay box for this), and/or lighting up the LEDs on your wearable.
Location – We have placed router radios around the venue and these provide us with coverage so that our
wearable radios can connect to the Internet. We’d like these routers to execute a “node discovery”
command periodically, reporting to Pachube how many radios are nearby and what are their
IDs. This will give us approximate location information about mobile radios, because it will tell us that mobile radio X is near fixed radio Y.
Triggers- Pachube allows us to set “triggers” that execute based on values in datastreams. For instance,
if we have location information (previous paragraph) we could tell a sensor in the breakfast room
to tweet the number of participants in the room at 20 minute intervals. Triggers are represented
as URLs, so they have to point to simple program such as written in PHP or Phython.
Please contact team leaders below if you want to volunteer to participate on any of the teams.
See you tomorrow for day 3!