adam-sim 0.1.1

A simulator of ADAM (Autonomous Domestic Ambidextrous Manipulator), a mobile robot manipulator consisting of a base with two Degrees of Freedom (DoF) and two Universal Robots UR3 of 6 DoF each.

ADAM simulator

ADAM (Autonomous Domestic Ambidextrous Manipulator) is a mobile robot manipulator consisting of a base with two Degrees of Freedom (DoF) and two Universal Robots UR3 of 6 DoF each.

The simulation was built using MuJoCo, a free and open source physics engine designed from the ground up for the purpose of model-based optimization, and in particular optimization through contacts.

Installation

Follow the next steps for installing the simulation on your device.

Requierements:

• Ubuntu
• Python 3.10.0 or higher

Install miniconda (highly-recommended)

It is highly recommended to install all the dependencies on a new virtual environment. For more information check the conda documentation for installation and environment management. For creating the environment use the following commands on the terminal.

conda create -n adam python==3.10.8
conda activate adam

Install from pip

The ADAM simulator is available as a pip package. For installing it just use:

pip install adam-sim

Install from source

Firstly, clone the repository in your system.

git clone https://github.com/vistormu/adam_simulator.git

Then, enter the directory and install the required dependencies

cd adam_simulator
pip install -r requirements.txt

Minimal Working Examples

In this example, ADAM moves the left and right wrist_2 continuously.

from adam import Simulation
from adam.entities import Configuration, Data


def main():
    sim: Simulation = Simulation()
    initial_data: Data = sim.load_scene()

    left_configuration: Configuration = initial_data.configuration.left_manipulator
    right_configuration: Configuration = initial_data.configuration.right_manipulator

    while sim.is_alive:
        left_configuration += Configuration(0.0, 0.0, 0.0, 0.0, 0.1, 0.0)
        right_configuration -= Configuration(0.0, 0.0, 0.0, 0.0, 0.1, 0.0)

        sim.render()
        data: Data = sim.step(left_configuration, right_configuration)

    sim.close()


if __name__ == '__main__':
    main()

In this example, a list of configurations are loaded and tested on the left manipulator.

from adam import Simulation, ConfigurationsManager
from adam.entities import Configuration, Data


def main():
    sim: Simulation = Simulation()
    initial_data: Data = sim.load_scene()

    configuration_list: list[Configuration] = ConfigurationsManager.load('test')

    for configuration in configuration_list:
        sim.render()

        data: Data = sim.step(configuration, initial_data.configuration.right_manipulator)

    sim.close()


if __name__ == '__main__':
    main()

API Reference

The developed API aims to wrap the MuJoCo API in order to simply move the manipulators of ADAM. The API is object-oriented and gives type annotations.

NOTE: the documentation is still in progress.

Data class

The Data class contains all the information related to the state of the manipulators. Currently the information offered is of the current configuration and the collision info.

class Data:
    configuration: ConfigurationData
    collision: CollisionData

Configuration Data class

The ConfigurationData class contains the configuration data of the left are right manipulator separately.

class ConfigurationData:
    left_manipulator: Configuration
    right_manipulator: Configuration

Configuration class

The Configuration class contains the information of a configuration of the manipulator.

class Configuration(NamedTuple):
    q1: float
    q2: float
    q3: float
    q4: float
    q5: float
    q6: float

    def to_degrees(self) -> Configuration
    def to_radians(self) -> Configuration
    def to_numpy(self) -> np.ndarray

Collision Data class

The CollisionData class contains the collision data of the left and right manipulator separately.

class CollisionData:
    left_manipulator: Collision
    right_manipulator: Collision

Collision class

The Collision class contains all the information of the collision of a manipulator.

class Collision(NamedTuple):
    collided: bool
    vector: list[bool]
    shoulder: list[str]
    upper_arm: list[str]
    forearm: list[str]
    wrist_1: list[str]
    wrist_2: list[str]
    wrist_3: list[str]

Simulation class

The Simulation class offers all the methods related to the simulation.

class Simulation:
    is_alive: bool

    def load_scene(self, scene: str) -> Data
    def set_view(self, center: tuple[float, float, float], azimuth: float, elevation: float, distance: float) -> None
    def extend_collisions(collision_dict: dict[int, str]) -> None:
    def step(self, left_configuration: Configuration, right_configuration: Configuration) -> Data
    def render(self, *, fps: int = 15) -> None
    def close(self) -> None

Configurations Manager class

The ConfigurationsManager class loads and saves a list of Configuration from a .csv file from a given directory.

class ConfigurationsManager:
    @staticmethod
    def load(filename: str) -> list[Configuration]

    @staticmethod
    def save(filename: str, configuration_list: list[Configuration]) -> None