Text exchange with a friend about how to defend crowds from threats like the shooter in Las Vegas.
Yup, saving the world, or trying to, one bear at a time.
Triangulating on a sound with data from thousands of willing opt-in smart phones is possible. Pitch, yaw, acceleration, relative volume compared to those in proximity to normalize. Calculate position from last known good if towers go out.
Mesh grid relative to each other if no service. Share UDP 5353 and change multicast DNS into a “people finder”.
The app, when turned on, would send a cascade of data flowing in with lots of noise. The analysis is the same thing anyone who has done log analysis with an ELK stack is familiar with. Have a buffer of say 10 seconds backwards until triggered.
With a few datasets from simulations (like having 30 people in a room and see if the app can figure out who blew the dog whistle.
Sensor based smart phone triangulation is one way we could defend ourselves in an attack on any soft target.
Note: the concept is somewhat related to what we are building at somarobotics.com. However I’m putting it out there because I’d love to see someone build a system to automatically respond and help.