2026 CS3 RET debut

CS3 AI
Teaching Kit

A hybrid platform for teaching artificial intelligence through browser lessons, edge hardware, physical sensors, and agentic coding.

Open the online toolkit Get the offline kit

Open CS3 AI Teaching Kit showing a Jetson Orin Nano Super, camera, sensor board, display, and keyboard

50+lectures across both components

30sensors for physical computing

17interactive browser lessons

38offline hardware lessons

Online and offline learning

One teaching system, from neural networks to edge AI.

The 2026 CS3 RET program debuts the CS3 AI Teaching Kit. Built around the NVIDIA Jetson Orin Nano Super, this hybrid kit combines online and offline components, 30 sensors, more than 50 lectures, and agentic AI integration.

The kit features demonstrations that teach computer vision through traffic counting, segmentation, parking assistance, and related projects. Agentic coding workflows help educators inspect, adapt, and extend each activity.

The online browser application includes 17 lessons, beginning with an introduction to neural networks and ending with large language models and how they are trained.

Two complementary components

Teach in the browser. Build at the edge.

Online component

AI Training Toolkit

Seventeen guided lessons introduce neural networks, image models, language models, and the practical ideas behind training. Everything runs in a browser.

Launch the online toolkit

Offline component

Jetson Teaching Kit

Thirty-eight hardware lessons and ten advanced programs connect cameras, sensors, edge inference, and agentic coding on the Jetson Orin Nano Super.

Browse the Hugging Face dataset Ubuntu installation

Built with educators

From demonstration to classroom project.

Teachers work directly with the hardware, test computer vision applications, and use coding agents to understand and reshape each project.

Teachers collaborating around CS3 teaching kits and laptops during a workshop — Collaborative, project-based learning

CS3 researchers and teachers examining a sensor and camera demonstration — Hardware and sensor exploration

Live facial-expression computer vision demonstration running on the teaching kit — Computer vision running at the edge

What educators can explore

A platform for seeing, sensing, and building.

Computer vision

Traffic counting, detection, segmentation, face tracking, parking assistance, depth, and pose projects.

Physical sensing

Thirty sensors connect AI concepts to motion, distance, light, sound, temperature, displays, and controls.

Agentic coding

Guided workflows help teachers read unfamiliar code, test changes, diagnose errors, and create extensions.

Offline instruction

Self-contained HTML decks, PDFs, notes, programs, models, and media support teaching without a network.

Ubuntu setup

Install or refresh the offline kit.

Run these commands on the Jetson or another Ubuntu computer. The final command downloads the current dataset to the Desktop and overwrites packaged files with the newest versions.

Local files that are not part of the published kit are left in place. Close any open lesson programs before updating.

Open the complete installation README

Terminal

sudo apt update
sudo apt install -y python3-venv

python3 -m venv "$HOME/.venvs/huggingface"
"$HOME/.venvs/huggingface/bin/pip" install \
  --upgrade huggingface_hub

"$HOME/.venvs/huggingface/bin/hf" download \
  mehmetkeremturkcan/cs3-ai-teaching-kit \
  --repo-type dataset \
  --local-dir "$HOME/Desktop/cs3-teaching-kit" \
  --force-download

Start with the component that fits your classroom.

Use the browser toolkit immediately, or download the full offline package for Jetson-based projects.

Online toolkit Offline dataset

CS3 AITeaching Kit