Array of Things Public Meeting Notes
October 18, 2017 | 5:30pm — 7:30pm | Association House
Example of what notes become once they are organized and cleaned.
The website has not changed and stated a lot expectations that have passed. Is there a commitment to show that the project is alive? Can you give us some background why the nodes did not go out as expected?
- CC: Came to this project as computer scientists, and put up 12 nodes in August and September of last year, but focused on how to make this reliable before putting more nodes up. Delay due to finding things that needed to fix, and spent time trying to fix things– water was getting into the device, moved from building them in the Argonne Laboratory to being manufactured offsite. Pushed too aggressively and faster than.
- CC: Where are we now? Learned not to the point to the fence too soon or committing to a schedule too soon. Confident that can get next 100 up and they will be reliable.
- CC: We can do a better job at updating the website.
- Questioner: it looked like people became apathetic. Questioner recommends updating the website on a more regular basis and share when things go wrong. (participated in privacy meeting)
What is the range of each individual sensor?
- CC: Temperature is a few miles… some nodes are places closer together, and others are further apart. This helps provide data on how nodes should be placed in the future
Could we optimize the box by reducing the complexity of the box and moving that into a data center?
- CC: Move programming out would have simplified the project. Wants device to be autonomous
- CC: Doing a project in rural India to build a $100 device to capture temperature, humidity, air quality
Are you measuring light?
- CC: Measuring overall light and infrared light. Measuring temperature, air quality, etc
- CC: Would like the measure methane of carbon dioxide but the price “isn’t right” for the sensors
Is there thought going into upgradability? Or create a stackable structure with new technologies/features?
- CC: Learned that by working with the electricians that it is easier to do a swap of the technology vs opening it up and changing things while they are up. Strategy for the upgrades is to make changes in lab and then make it efficient for CDOT electricians to switch it
- CC: Can add layers to make it taller and add new features there.
What type of internet infrastructure is required? Is that equally distributed throughout the city?
- CC: Cellular at the top of the dome and partnered with AT&T to do all of the data communications. Some are connected with wifi and others with an ethernet cable. Difficult for the wifi to be reliable and secure. Every 30 seconds we capture the data.
- CC: When we will need a faster speed is when the software needs to be updated.
Will there be a point that sensor data to affect how the city reacts to the data? Ex: if there are sirens could it switch the traffic lights.
- CC: Providing data that is recent (within a minute or so) so interested parties can build those capabilities. Ex: flooding data could connect to 311
- Example of Louisville (missed the exact example)
- Access to open data as research to affect the local environment. Open data to create new solutions
Sensors in the camera for ordering 311 services?
- DD: Flooding is an example use case; possibly potholes if the camera can see it and that orders a 311 services. That would happen over time working with researchers
- CC: City of Portland bike share program that would encourage if people are riding without helmets.
Sound sensors to capture gunshots and where?
- DD: use case to capture gunshots and immediately call Chicago Police Department
- CC: Sound and kinds of sound in neighborhoods to determine things like why you might have insomnia. Purpose is to support that type of research
- CC: possible use case is having a sound sensor where a specific whistle sound could alert campus security of danger
Positive comment from an educator of Humboldt Park.
Do you process part of the data and then make it available? Or is it raw data?
- CC: It’s a lot of data, but for the University of Chicago it’s not a data to store.
- CC: Plenar.io has some data streams
How do you ensure your sensors are working?
- CC: There are eight sites that are run by Cook County Environmental Protection Agency who are taking air quality readings and there is a plan to put 8 nodes at those sites to compare.
- CC: Ability to pick out data from a particular node in your community, and that would be a transformation that happens at Argonne
How do you ensure that data will be comparable when systems are upgraded?
- CC: This is a challenge with networks like these, information will be stored in the metadata.
How long has this project been going on?
- CC: Idea came about 5 years ago, and received grant 2 years ago to do the 500 nodes. In 2018 we will be “commissioning the instrument” (like closing the house)
Can you speak more to the vibrations and how those used?
- CC: accelerometer + magnetometer (sp?) cost about $6
- Could tell about an earthquake, or something like a truck coming down the street. Not sure what it could do but it was cheap enough to test it
Availability of data? There is data from two nodes but where is it accessed?
- Data can pulled out by programmers, but there isn’t a public facing set of data until 2018
- Data that is not published is either being considered misleading
- Data can be extracted by a specific node to get a value out, but learned that developers may have wanted something different
- Hope is to leverage the OpenGrid.io website to be able to see where nodes are and to click the node to get the data
- DD: if you have ideas as to how you want to get the data and then this could inform changes with OpenGrid
Community brainstorming (30 min.)
Note: Table worksheets will be collected and pictures will be taken of maps and posters around the room. No notes are needed!