This environment is part of the MAgent environments. Please read that page first for general information.


from pettingzoo.sisl import multiwalker_v9



Parallel API


Manual Control



agents= ['walker_0', 'walker_1', 'walker_2']



Action Shape


Action Values

(-1, 1)

Observation Shape


Observation Values


Average Total Reward



In this environment, bipedal robots attempt to carry a package placed on top of them towards the right. By default, the number of robots is set to 3.

Each walker receives a reward equal to the change in position of the package from the previous timestep, multiplied by the forward_reward scaling factor. The maximum achievable total reward depends on the terrain length; as a reference, for a terrain length of 75, the total reward under an optimal policy is around 300.

The environment is done if the package falls, or if the package goes beyond the left edge of the terrain. By default, the environment is also done if any walker falls. In all of these cases, each walker receives a reward of -100. The environment is also done if package falls off the right edge of the terrain, with reward 0.

When a walker falls, it receives an additional penalty of -10. If terminate_on_fall = False, then the environment is not done when the walker falls, but only when the package falls. If remove_on_fall = True, the fallen walker is removed from the environment. The agents also receive a small shaped reward of -5 times the change in their head angle to keep their head oriented horizontally.

If shared_reward is chosen (True by default), the agents’ individual rewards are averaged to give a single mean reward, which is returned to all agents.

Each walker exerts force on two joints in their two legs, giving a continuous action space represented as a 4 element vector. Each walker observes via a 31 element vector containing simulated noisy lidar data about the environment and information about neighboring walkers. The environment’s duration is capped at 500 frames by default (can be controlled by the max_cycles setting).

Observation Space#

Each agent receives an observation composed of various physical properties of its legs and joints, as well as LIDAR readings from the space immediately in front and below the robot. The observation also includes information about neighboring walkers, and the package. The neighbour and package observations have normally distributed signal noise controlled by position_noise and angle_noise. For walkers without neighbors, observations about neighbor positions are zero.

This table enumerates the observation space:

Index: [start, end)




Hull angle

[0, 2*pi]


Hull angular velocity

[-inf, inf]


X Velocity

[-1, 1]


Y Velocity

[-1, 1]


Hip joint 1 angle

[-inf, inf]


Hip joint 1 speed

[-inf, inf]


Knee joint 1 angle

[-inf, inf]


Knee joint 1 speed

[-inf, inf]


Leg 1 ground contact flag

{0, 1}


Hip joint 1 angle

[-inf, inf]


Hip joint 2 speed

[-inf, inf]


Knee joint 2 angle

[-inf, inf]


Knee joint 2 speed

[-inf, inf]


Leg 2 ground contact flag

{0, 1}


LIDAR sensor readings

[-inf, inf]


Left neighbor relative X position (0.0 for leftmost walker) (Noisy)

[-inf, inf]


Left neighbor relative Y position (0.0 for leftmost walker) (Noisy)

[-inf, inf]


Right neighbor relative X position (0.0 for rightmost walker) (Noisy)

[-inf, inf]


Right neighbor relative Y position (0.0 for rightmost walker) (Noisy)

[-inf, inf]


Walker X position relative to package (0 for left edge, 1 for right edge) (Noisy)

[-inf, inf]


Walker Y position relative to package (Noisy)

[-inf, inf]


Package angle (Noisy)

[-inf, inf]


multiwalker_v9.env(n_walkers=3, position_noise=1e-3, angle_noise=1e-3, forward_reward=1.0, terminate_reward=-100.0, fall_reward=-10.0, shared_reward=True,
terminate_on_fall=True, remove_on_fall=True, terrain_legth=200, max_cycles=500)

n_walkers: number of bipedal walker agents in environment

position_noise: noise applied to neighbors and package positional observations

angle_noise: noise applied to neighbors and package rotational observations

forward_reward: reward received is forward_reward * change in position of the package

fall_reward: reward applied when an agent falls

shared_reward: whether reward is distributed among all agents or allocated individually

terminate_reward: reward applied to each walker if they fail to carry the package to the right edge of the terrain

terminate_on_fall: If True (default), a single walker falling causes all agents to be done, and they all receive an additional terminate_reward. If False, then only the fallen agent(s) receive fall_reward, and the rest of the agents are not done i.e. the environment continues.

remove_on_fall: Remove a walker when it falls (only works when terminate_on_fall is False)

terrain_length: length of terrain in number of steps

max_cycles: after max_cycles steps all agents will return done

Version History#

  • v8: Replaced local_ratio, fixed rewards, terrain length as an argument and documentation (1.15.0)

  • v7: Fixed problem with walker collisions (1.8.2)

  • v6: Fixed observation space and made large improvements to code quality (1.5.0)

  • v5: Fixes to reward structure, added arguments (1.4.2)

  • v4: Misc bug fixes (1.4.0)

  • v3: Fixes to observation space (1.3.3)

  • v2: Various fixes and environment argument changes (1.3.1)

  • v1: Fixes to how all environments handle premature death (1.3.0)

  • v0: Initial versions release (1.0.0)