import os
from typing import Optional
import mujoco_py.builder
import numpy as np
from gym import utils
from gym.envs.mujoco import MujocoEnv
# XML Variables
ROBOT_COLLISION_OBJ = ["wrist_palm_link_convex_geom",
"wrist_pitch_link_convex_decomposition_p1_geom",
"wrist_pitch_link_convex_decomposition_p2_geom",
"wrist_pitch_link_convex_decomposition_p3_geom",
"wrist_yaw_link_convex_decomposition_p1_geom",
"wrist_yaw_link_convex_decomposition_p2_geom",
"forearm_link_convex_decomposition_p1_geom",
"forearm_link_convex_decomposition_p2_geom",
"upper_arm_link_convex_decomposition_p1_geom",
"upper_arm_link_convex_decomposition_p2_geom",
"shoulder_link_convex_decomposition_p1_geom",
"shoulder_link_convex_decomposition_p2_geom",
"shoulder_link_convex_decomposition_p3_geom",
"base_link_convex_geom", "table_contact_geom"]
CUP_COLLISION_OBJ = ["cup_geom_table3", "cup_geom_table4", "cup_geom_table5", "cup_geom_table6",
"cup_geom_table7", "cup_geom_table8", "cup_geom_table9", "cup_geom_table10",
"cup_geom_table15", "cup_geom_table16", "cup_geom_table17", "cup_geom1_table8"]
class BeerPongEnv(MujocoEnv, utils.EzPickle):
def __init__(self, frame_skip=2):
self._steps = 0
# Small Context -> Easier. Todo: Should we do different versions?
# self.xml_path = os.path.join(os.path.dirname(os.path.abspath(__file__)), "assets",
# "beerpong_wo_cup" + ".xml")
# self._cup_pos_min = np.array([-0.32, -2.2])
# self._cup_pos_max = np.array([0.32, -1.2])
self.xml_path = os.path.join(os.path.dirname(os.path.abspath(__file__)), "assets",
"beerpong_wo_cup_big_table" + ".xml")
self._cup_pos_min = np.array([-1.42, -4.05])
self._cup_pos_max = np.array([1.42, -1.25])
self._start_pos = np.array([0.0, 1.35, 0.0, 1.18, 0.0, -0.786, -1.59])
self._start_vel = np.zeros(7)
self.release_step = 100 # time step of ball release
self.ep_length = 600 // frame_skip
self.repeat_action = frame_skip
# TODO: If accessing IDs is easier in the (new) official mujoco bindings, remove this
self.model = None
self.site_id = lambda x: self.model.site_name2id(x)
self.body_id = lambda x: self.model.body_name2id(x)
self.geom_id = lambda x: self.model.geom_name2id(x)
# for reward calculation
self.dists = []
self.dists_final = []
self.action_costs = []
self.ball_ground_contact_first = False
self.ball_table_contact = False
self.ball_wall_contact = False
self.ball_cup_contact = False
self.ball_in_cup = False
self.dist_ground_cup = -1 # distance floor to cup if first floor contact
MujocoEnv.__init__(self, self.xml_path, frame_skip=1, mujoco_bindings="mujoco_py")
utils.EzPickle.__init__(self)
@property
def start_pos(self):
return self._start_pos
@property
def start_vel(self):
return self._start_vel
def reset(self, *, seed: Optional[int] = None, return_info: bool = False, options: Optional[dict] = None):
self.dists = []
self.dists_final = []
self.action_costs = []
self.ball_ground_contact_first = False
self.ball_table_contact = False
self.ball_wall_contact = False
self.ball_cup_contact = False
self.ball_in_cup = False
self.dist_ground_cup = -1 # distance floor to cup if first floor contact
return super().reset()
def reset_model(self):
init_pos_all = self.init_qpos.copy()
init_pos_robot = self.start_pos
init_vel = np.zeros_like(init_pos_all)
self._steps = 0
start_pos = init_pos_all
start_pos[0:7] = init_pos_robot
# TODO: Ask Max why we need to set the state twice.
self.set_state(start_pos, init_vel)
start_pos[7::] = self.data.site_xpos[self.site_id("init_ball_pos"), :].copy()
self.set_state(start_pos, init_vel)
xy = self.np_random.uniform(self._cup_pos_min, self._cup_pos_max)
xyz = np.zeros(3)
xyz[:2] = xy
xyz[-1] = 0.840
self.model.body_pos[self.body_id("cup_table")] = xyz
return self._get_obs()
def step(self, a):
crash = False
for _ in range(self.repeat_action):
applied_action = a + self.data.qfrc_bias[:len(a)].copy() / self.model.actuator_gear[:, 0]
try:
self.do_simulation(applied_action, self.frame_skip)
# self.reward_function.check_contacts(self.sim) # I assume this is not important?
if self._steps < self.release_step:
self.data.qpos[7::] = self.data.site_xpos[self.site_id("init_ball_pos"), :].copy()
self.data.qvel[7::] = self.data.site_xvelp[self.site_id("init_ball_pos"), :].copy()
crash = False
except mujoco_py.builder.MujocoException:
crash = True
ob = self._get_obs()
if not crash:
reward, reward_infos = self._get_reward(applied_action)
is_collided = reward_infos['is_collided']
done = is_collided or self._steps == self.ep_length - 1
self._steps += 1
else:
reward = -30
done = True
reward_infos = {"success": False, "ball_pos": np.zeros(3), "ball_vel": np.zeros(3), "is_collided": False}
infos = dict(
reward=reward,
action=a,
q_pos=self.data.qpos[0:7].ravel().copy(),
q_vel=self.data.qvel[0:7].ravel().copy(), sim_crash=crash,
)
infos.update(reward_infos)
return ob, reward, done, infos
def _get_obs(self):
theta = self.data.qpos.flat[:7]
theta_dot = self.data.qvel.flat[:7]
ball_pos = self.data.get_body_xpos("ball").copy()
cup_goal_diff_final = ball_pos - self.data.get_site_xpos("cup_goal_final_table").copy()
cup_goal_diff_top = ball_pos - self.data.get_site_xpos("cup_goal_table").copy()
return np.concatenate([
np.cos(theta),
np.sin(theta),
theta_dot,
cup_goal_diff_final,
cup_goal_diff_top,
self.model.body_pos[self.body_id("cup_table")][:2].copy(),
# [self._steps], # Use TimeAwareObservation Wrapper instead ....
])
@property
def dt(self):
return super(BeerPongEnv, self).dt * self.repeat_action
def _get_reward(self, action):
goal_pos = self.data.get_site_xpos("cup_goal_table")
ball_pos = self.data.get_body_xpos("ball")
ball_vel = self.data.get_body_xvelp("ball")
goal_final_pos = self.data.get_site_xpos("cup_goal_final_table")
self._check_contacts()
self.dists.append(np.linalg.norm(goal_pos - ball_pos))
self.dists_final.append(np.linalg.norm(goal_final_pos - ball_pos))
self.dist_ground_cup = np.linalg.norm(ball_pos - goal_pos) \
if self.ball_ground_contact_first and self.dist_ground_cup == -1 else self.dist_ground_cup
action_cost = np.sum(np.square(action))
self.action_costs.append(np.copy(action_cost))
# # ##################### Reward function which does not force to bounce once on the table (quad dist) #########
# Is this needed?
# self._is_collided = self._check_collision_with_itself([self.geom_id(name) for name in CUP_COLLISION_OBJ])
if self._steps == self.ep_length - 1: # or self._is_collided:
min_dist = np.min(self.dists)
final_dist = self.dists_final[-1]
if self.ball_ground_contact_first:
min_dist_coeff, final_dist_coeff, ground_contact_dist_coeff, rew_offset = 1, 0.5, 2, -4
else:
if not self.ball_in_cup:
if not self.ball_table_contact and not self.ball_cup_contact and not self.ball_wall_contact:
min_dist_coeff, final_dist_coeff, ground_contact_dist_coeff, rew_offset = 1, 0.5, 0, -4
else:
min_dist_coeff, final_dist_coeff, ground_contact_dist_coeff, rew_offset = 1, 0.5, 0, -2
else:
min_dist_coeff, final_dist_coeff, ground_contact_dist_coeff, rew_offset = 0, 1, 0, 0
action_cost = 1e-4 * np.mean(action_cost)
reward = rew_offset - min_dist_coeff * min_dist ** 2 - final_dist_coeff * final_dist ** 2 - \
action_cost - ground_contact_dist_coeff * self.dist_ground_cup ** 2
# release step punishment
min_time_bound = 0.1
max_time_bound = 1.0
release_time = self.release_step * self.dt
release_time_rew = int(release_time < min_time_bound) * (-30 - 10 * (release_time - min_time_bound) ** 2) + \
int(release_time > max_time_bound) * (-30 - 10 * (release_time - max_time_bound) ** 2)
reward += release_time_rew
success = self.ball_in_cup
else:
action_cost = 1e-2 * action_cost
reward = - action_cost
success = False
# ##############################################################################################################
infos = {"success": success, "ball_pos": ball_pos.copy(),
"ball_vel": ball_vel.copy(), "action_cost": action_cost, "task_reward": reward,
"table_contact_first": int(not self.ball_ground_contact_first),
"is_collided": False} # TODO: Check if is collided is needed
return reward, infos
def _check_contacts(self):
if not self.ball_table_contact:
self.ball_table_contact = self._check_collision({self.geom_id("ball_geom")},
{self.geom_id("table_contact_geom")})
if not self.ball_cup_contact:
self.ball_cup_contact = self._check_collision({self.geom_id("ball_geom")},
{self.geom_id(name) for name in CUP_COLLISION_OBJ})
if not self.ball_wall_contact:
self.ball_wall_contact = self._check_collision({self.geom_id("ball_geom")},
{self.geom_id("wall")})
if not self.ball_in_cup:
self.ball_in_cup = self._check_collision({self.geom_id("ball_geom")},
{self.geom_id("cup_base_table_contact")})
if not self.ball_ground_contact_first:
if not self.ball_table_contact and not self.ball_cup_contact and not self.ball_wall_contact \
and not self.ball_in_cup:
self.ball_ground_contact_first = self._check_collision({self.geom_id("ball_geom")},
{self.geom_id("ground")})
# Checks if id_set1 has a collision with id_set2
def _check_collision(self, id_set_1, id_set_2):
"""
If id_set_2 is set to None, it will check for a collision with itself (id_set_1).
"""
collision_id_set = id_set_2 - id_set_1 if id_set_2 is not None else id_set_1
for coni in range(self.data.ncon):
con = self.data.contact[coni]
if ((con.geom1 in id_set_1 and con.geom2 in collision_id_set) or
(con.geom2 in id_set_1 and con.geom1 in collision_id_set)):
return True
return False
class BeerPongEnvFixedReleaseStep(BeerPongEnv):
def __init__(self, frame_skip=2):
super().__init__(frame_skip)
self.release_step = 62 # empirically evaluated for frame_skip=2!
class BeerPongEnvStepBasedEpisodicReward(BeerPongEnv):
def __init__(self, frame_skip=2):
super().__init__(frame_skip)
self.release_step = 62 # empirically evaluated for frame_skip=2!
def step(self, a):
if self._steps < self.release_step:
return super(BeerPongEnvStepBasedEpisodicReward, self).step(a)
else:
reward = 0
done = False
while not done:
sub_ob, sub_reward, done, sub_infos = super(BeerPongEnvStepBasedEpisodicReward, self).step(
np.zeros(a.shape))
reward += sub_reward
infos = sub_infos
ob = sub_ob
ob[-1] = self.release_step + 1 # Since we simulate until the end of the episode, PPO does not see the
# internal steps and thus, the observation also needs to be set correctly
return ob, reward, done, infos
# class ALRBeerBongEnvStepBased(ALRBeerBongEnv):
# def __init__(self, frame_skip=1, apply_gravity_comp=True, noisy=False, rndm_goal=False, cup_goal_pos=None):
# super().__init__(frame_skip, apply_gravity_comp, noisy, rndm_goal, cup_goal_pos)
# self.release_step = 62 # empirically evaluated for frame_skip=2!
#
# def step(self, a):
# if self._steps < self.release_step:
# return super(ALRBeerBongEnvStepBased, self).step(a)
# else:
# reward = 0
# done = False
# while not done:
# sub_ob, sub_reward, done, sub_infos = super(ALRBeerBongEnvStepBased, self).step(np.zeros(a.shape))
# if not done or sub_infos['sim_crash']:
# reward += sub_reward
# else:
# ball_pos = self.sim.data.body_xpos[self.sim.model._body_name2id["ball"]].copy()
# cup_goal_dist_final = np.linalg.norm(ball_pos - self.sim.data.site_xpos[
# self.sim.model._site_name2id["cup_goal_final_table"]].copy())
# cup_goal_dist_top = np.linalg.norm(ball_pos - self.sim.data.site_xpos[
# self.sim.model._site_name2id["cup_goal_table"]].copy())
# if sub_infos['success']:
# dist_rew = -cup_goal_dist_final ** 2
# else:
# dist_rew = -0.5 * cup_goal_dist_final ** 2 - cup_goal_dist_top ** 2
# reward = reward - sub_infos['action_cost'] + dist_rew
# infos = sub_infos
# ob = sub_ob
# ob[-1] = self.release_step + 1 # Since we simulate until the end of the episode, PPO does not see the
# # internal steps and thus, the observation also needs to be set correctly
# return ob, reward, done, infos
if __name__ == "__main__":
env = BeerPongEnv(frame_skip=2)
env.seed(0)
# env = ALRBeerBongEnvStepBased(frame_skip=2)
# env = ALRBeerBongEnvStepBasedEpisodicReward(frame_skip=2)
# env = ALRBeerBongEnvFixedReleaseStep(frame_skip=2)
import time
env.reset()
env.render("human")
for i in range(600):
# ac = 10 * env.action_space.sample()
ac = 0.05 * np.ones(7)
obs, rew, d, info = env.step(ac)
env.render("human")
if d:
print('reward:', rew)
print('RESETTING')
env.reset()
time.sleep(1)
env.close()