Implementation:ARISE Initiative Robosuite VisualizationWrapper

Metadata:

robosuite
Robotics_Simulation
Teleoperation
last_updated: 2026-02-15 12:00 GMT

Overview

Concrete wrapper for adding visual indicator sites to robosuite environments provided by the wrappers module.

Description

The `VisualizationWrapper` class wraps a `MujocoEnv` to add visual indicator sites (spheres or custom shapes) into the MuJoCo model. These indicators provide real-time visual feedback during teleoperation. Supports single default indicator, custom indicator configs, or multiple indicators.

The wrapper modifies the MuJoCo model XML at initialization time to inject visual geometry (sites) that can be dynamically positioned during runtime. This allows operators to see target positions, waypoints, or other spatial goals directly in the simulation rendering without cluttering the environment's core definition.

Usage

Wrap an environment when using teleoperation devices. Typically applied before DataCollectionWrapper in demo collection pipelines. The wrapper is transparent to the environment's core functionality and does not affect observation or action spaces.

Code Reference

Source: robosuite

File: robosuite/wrappers/visualization_wrapper.py

Lines: L23-186

Signature:

class VisualizationWrapper:
    def __init__(self, env, indicator_configs=None):
        """
        Args:
            env (MujocoEnv): The environment to visualize
            indicator_configs (None or str or dict or list): Configurations for visual indicators.
                None = no indicators, 'default' = spherical indicator, dict = single config, list = multiple configs
        """

Import:

from robosuite.wrappers import VisualizationWrapper

I/O Contract

Inputs:

env (MujocoEnv, Required): The base environment to wrap and augment with visual indicators
indicator_configs (None/str/dict/list, Optional): Configuration specifying the visual indicators to add
- `None` - No indicators added
- `'default'` - Adds a default spherical indicator site
- `dict` - Single custom indicator configuration (position, size, color, shape)
- `list` - Multiple indicator configurations for multiple visual markers

Outputs:

VisualizationWrapper instance: A wrapped environment that delegates all MujocoEnv methods (reset, step, render, etc.) and adds indicator sites to the MuJoCo model for visual feedback

Usage Examples

Example 1: Wrap with default indicator

import robosuite as suite
from robosuite.wrappers import VisualizationWrapper

# Create base environment
env = suite.make(
    env_name="Lift",
    robots="Panda",
    has_renderer=True,
    has_offscreen_renderer=False,
    use_camera_obs=False,
)

# Wrap with default indicator
env = VisualizationWrapper(env, indicator_configs='default')

# Reset environment (indicator is now visible in rendering)
obs = env.reset()

# Update indicator position during teleoperation
target_position = [0.5, 0.0, 0.8]
env.sim.data.set_mocap_pos('indicator_default', target_position)

# Step through environment
for _ in range(100):
    action = env.action_space.sample()
    obs, reward, done, info = env.step(action)
    env.render()
    if done:
        obs = env.reset()

Example 2: Wrap with custom indicator config

import robosuite as suite
from robosuite.wrappers import VisualizationWrapper

# Create base environment
env = suite.make(
    env_name="PickPlace",
    robots="Sawyer",
    has_renderer=True,
    has_offscreen_renderer=False,
    use_camera_obs=False,
)

# Define custom indicator configuration
custom_indicator = {
    'name': 'target_marker',
    'pos': [0.6, 0.1, 0.9],
    'size': [0.02, 0.02, 0.02],
    'rgba': [1.0, 0.0, 0.0, 0.5],  # Semi-transparent red
    'type': 'sphere'
}

# Wrap with custom indicator
env = VisualizationWrapper(env, indicator_configs=custom_indicator)

# Reset and use
obs = env.reset()

# Multiple indicators example
multiple_indicators = [
    {'name': 'waypoint1', 'pos': [0.5, 0.0, 0.8], 'rgba': [0.0, 1.0, 0.0, 0.5]},
    {'name': 'waypoint2', 'pos': [0.6, 0.2, 0.8], 'rgba': [0.0, 0.0, 1.0, 0.5]},
    {'name': 'goal', 'pos': [0.7, 0.0, 1.0], 'rgba': [1.0, 0.0, 0.0, 0.8]}
]

env_multi = suite.make("Stack", robots="Panda", has_renderer=True)
env_multi = VisualizationWrapper(env_multi, indicator_configs=multiple_indicators)

Related Pages

Page Connections

Double-click a node to navigate. Hold to expand connections.

Principle

Implementation

Heuristic

Environment