https://github.com/upkie/new_agent

Template to create new agents for Upkie
https://github.com/upkie/new_agent

legged-robots python robotics simulation wheeled-biped

Last synced: 30 days ago
JSON representation

Template to create new agents for Upkie

• Host: GitHub
• URL: https://github.com/upkie/new_agent
• Owner: upkie
• License: apache-2.0
• Created: 2023-10-24T15:05:39.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2024-10-14T05:36:45.000Z (4 months ago)
• Last Synced: 2024-11-11T01:18:33.859Z (3 months ago)
• Topics: legged-robots, python, robotics, simulation, wheeled-biped
• Language: C++
• Homepage:
• Size: 73.2 KB
• Stars: 0
• Watchers: 1
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# New agent template

[![upkie](https://img.shields.io/badge/upkie-6.0.0-salmon)](https://github.com/upkie/upkie/tree/v6.0.0)

Template repository to create new agents with custom spines for [Upkie](https://github.com/upkie/upkie) wheeled bipeds.

If you don't need a custom spine, you can implement your agent in Python directly, as done for instance in the [MPC](https://github.com/upkie/mpc_balancer) and [Pink](https://github.com/upkie/pink_balancer) balancers.

## Getting started

1. Create a new repository from this template

2. Search for the string "TODO" and update files accordingly

3. Replace `LICENSE` with the license of your choice (the default one is Apache-2.0)

4. Start listing your dependencies in `environment.yaml`

5. Rename and start implement your agent from `agent.py`

6. Implement your C++ spines in the `spines` directory

## Usage

- Install Python packages to a conda environment: `conda env create -f environment.yaml`

- Activate conda environment: `conda activate `

- Run the simulation spine: `make run_bullet_spine`

- Build the pi3hat spine locally: `make build`

- Upload the full repository (with built spines) to the robot: `make upload`

- Run the pi3hat spine: `make run_pi3hat_spine` (on robot)

- Run your agent: `python agent.py`