Soldiers and first responders do not have rapidly deployable
or scalable autonomous sensor support to use during recon, ambush, or
search-and-rescue operations. The reasons for lack of scalable, autonomous sensor
support to deployed soldiers are fourfold:
- A handful of sensors can overwhelm local network
capacity between the soldier or first responder, the team, and the devices that
are aiding them.
- There is no effective autonomy framework that
allows sensors and drones to operate effectively without a human in the loop.
- Squads in proximity may impede each other's
networks (i.e., scalability of current technology is lacking).
- There are concerns over efficiency, size, and
durability of drones and sensors in the hands of soldiers and first responders.
We aim to make inroads into each of these problem
areas by creating a scalable middleware framework that supports autonomous AI
for a search-and-rescue mission. The goal of this project is to demonstrate an
autonomous swarm of five thermal-enabled drones that search for survivors in a
field and report thermal targets back to a smartphone.
For related information, listen to the podcast "Human-in-the-Loop Autonomy" or watch the video "Multi-UAV Search and Rescue."