BulletArm: Further work on this package has been migrated to BulletArm.

Helping Hands RL Environments

• License

This package contains various simulated robotics environments used for research in the Helping Hands lab. The majority of these environments entail a robotic arm armed with a paralel jaw gripper executing a series of manipulation based tasks. For a full list of the tasks currently implemented see below. The core simulator used for most tasks is PyBullet but a simple numpy based simulator is included for quick prototyping.

Table of Contents

1. Requirments
2. Installation
3. Environments
4. Parameters
5. Publications

Requirments

Install

1. Install Python 3.7
2. Clone this repo

   git clone https://github.com/ColinKohler/helping_hands_rl_envs.git
   cd helping_hands_rl_envs
   

3. Install dependencies

   pip install -r requirements.txt 
   

4. Install this package

   pip install .
   

5. Run the block stacking demo

   import numpy as np
   from helping_hands_rl_envs import env_factory
   
   workspace = np.asarray([[0.3, 0.6],
                           [-0.15, 0.15],
                           [0, 0.50]])
   # environment parameters
   env_config = {'workspace': workspace, 'max_steps': 10, 'obs_size': 90, 'render': True, 'fast_mode': True,
                 'seed': 0, 'action_sequence': 'pxyr', 'num_objects': 4, 'random_orientation': True,
                 'reward_type': 'sparse', 'robot': 'kuka', 'workspace_check': 'point'}
   # planner parameters
   planner_config = {'random_orientation': True}
   # create 1 block_stacking environment
   env = env_factory.createEnvs(1, 'pybullet', 'block_stacking', env_config, planner_config)
   env.reset()
   for i in range(5, -1, -1):
       action = env.getNextAction()
       (states_, in_hands_, obs_), rewards, dones = env.step(action, auto_reset=False)
       input('press enter to continue')
   env.close()

Environments

2.5D Environments

• block_picking: Pick up all blocks.
• block_stacking: Pick up blocks and place them in a stack.
• block_adjacent: TODO
• brick_stacking: TODO
• pyramid_stacking: TODO
• house_building_1: Build a house structure with a block stack and a small triangle roof on top.
• house_building_2: Build a house structure with two block bases and a large triangle roof.
• house_building_3: Build a house structure with two block bases, one brick in the middle, and a large triangle roof.
• house_building_4: Build a house structure with two block bases, one brick in the second layer, two blocks in the third layer, and a large triangle roof.
• house_building_x: TODO
• improvise_house_building_2: Build a house structure with two random shape bases and a large triangle roof.
• improvise_house_building_3: Build a house structure with two random shape bases, one brick in the middle, and a large triangle roof.
• improvise_house_building_discrete: Build a house structure with 4 random shapes and a large triangle roof. The heights of the random shapes are sampled from two discrete numbers.
• improvise_house_building_random: Build a house structure with 4 random shapes and a large triangle roof. The heights of the random shapes are sampled from continuous numbers.

6D Environments

• ramp_block_stacking: Finish block_stacking in the workspace with two ramps
• ramp_house_building_1: Finish house_building_1 in the workspace with two ramps
• ramp_house_building_2: Finish house_building_2 in the workspace with two ramps
• ramp_house_building_3: Finish house_building_3 in the workspace with two ramps
• ramp_house_building_4: Finish house_building_4 in the workspace with two ramps
• ramp_improvise_house_building_2: Finish improvise_house_building_2 in the workspace with two ramps
• ramp_improvise_house_building_3: Finish improvise_house_building_3 in the workspace with two ramps

Parameters

Publications

If you like this package and use it in your own work, please cite this repository.