First Attempt at Modeling
Nov 15, 2019
On our journey to teach a drone to move around a room, we taught it to go around and look around, collecting images.
We then collected a few images and used them to instruct a model with TensorFlow.
Nov 12, 2019
Let’s start training our drone to explore the environment in search of objects.
We choose a very simple object to search for, in this case a red ball. And we use an equally simple movement: turn around.
Using Jupyter to Control the Drone
Nov 9, 2019
For our DroneTrainer project, it is useful to be able to use the most widespread tool to create DeepLearning applications: the Jupyter notebook.
For this reason we have modified the SumoControl control program so that it can be used by Jupyter.
Web Interface to the Drone
Nov 8, 2019
The Jumping Sumo drone manifacturer distributes an app that allows you to control it from your mobile, but we instead we prefer a web interface to it.
For our project, we are interested in controlling it with an artificial intelligence that will be executed in our Jetson Nano edge server.
Connecting to the Drone
Nov 7, 2019
The first challenge was to connect the edge server, based on the Jetson Nano, to the Jumping Sumo drone. The server uses a variant of the Linux Ubuntu operating system. The drone uses the ROS operating system which is also based on Linux.