DroneAid currently:
- Takes the video stream of a Tello drone and counts the number of DroneAid Symbols found.
- Contains a dashboard to fly a Tello drone.
- Works with a webcam.
Within the next months we are aiming to do:
- Complete the mapping component that will take the GPS data from reported images and plot them on map.
- Consume the pre-trained model as a Docker image.
- Extend to consumer DJI drones like the Mavic, Phantom, and Spark.
- Document steps for PowerAI and Watson Visual Recognition.
- Create a model to detect handrawn "SOS".
- Integrate reporting with the Project OWL mesh network.
Longer term goals are:
- Improve the accuracy of the model by using realistic simulated enviornments (floods, earthquakes, forest fires, etc).
- Produce a basic signaling kit with "SOS" on one side and "OK" on the other.
- Create a Twitter bot that will look for tweets with a hashtag (i.e. #DroneAidSOS) for social media reporting.
- Create a custom drone that will have DroneAid embedded inside, which will not require a computer or mobile app.
- Create a reporting dashboard for first responders and non-profit relief organizations.
