Dmitri Villevald
Published © GPL3+

Transfer Learning with NVIDIA JetBot - Fun with Cones

Teach your robot to find a path in a maze of traffic cones using the camera and state-of-the-art deep learning model.

IntermediateFull instructions provided10 hours3,622

Things used in this project

Hardware components

NVIDIA JetBot Wiki's Bill of Materials page lists everything you need to build JetBot, along with purchasing links from popular vendors.
Computer with NVIDIA GPU
Needed to train the model
BlueDot Trading 4” RC Racing Agility Cones, Orange – Set of 20

Hand tools and fabrication machines

Soldering iron (generic)
Soldering iron (generic)


Read more


Main update loop

Explains the main update loop and shows how robot action depends on the model predictions and the robot's current state


Three Jupyter notebooks for this project

The Jupyter notebooks data_collection_cones.ipynb, train_model_cones.ipynb and live_demo_cones.ipynb are located in subfolder traffic_cones_driving


Dmitri Villevald

Dmitri Villevald

5 projects • 18 followers
Manager, Advanced Analytics with a passion for Artificial Intelligence products.