This is a project by students of Northern New Mexico College (NNMC). The purpose of this project is to program a JetBot to interact with its environment by leveraging artificial intelligence (AI). We are in collaboration with Students & Teachers Assisting Real Research (STARR).
We have assembled two JetBots and programmed them to perform tasks such as data collection, navigation, object tracking, and object recognition. We continue to build on that work to achieve greater functionality.
Our JetBots were built using Sparkfun JetBot Kits: https://www.sparkfun.com/products/16417.
Each Sparkfun Kit was donated by a nonprofit group, the Computer Science Alliance in Santa Fe, New Mexico: https://sharenm.org/computer-science-alliance/computer-science-alliance.
The JetBots run on an Nvidia Jetson Nano, an AI development board featuring a Tegra System on a Chip (SoC) with a 128-core GPU. The JetBot OS is Linux for Tegra (L4T). The JetBot is programmed and managed with Python through Jupyter Notebooks. We are relying on PyTorch, a machine learning library, to train the JetBot.
The Python programming code originates from Nvidia's JetBot project here: https://github.com/NVIDIA-AI-IOT/jetbot
George Gorospe of NASA and STARR has modified the data collection notebook here: https://github.com/RFisherIves/STARR
We have also utilized code from DVillevald's JetBot project here: https://github.com/dvillevald/Finding-path-in-maze-of-traffic-cones/