BenchNPIN: Benchmarking Non-prehensile Interactive Navigation

BenchNPIN is a comprehensive suite of benchmarking tools for mobile robot non-prehensile interactive navigation. The goal of BenchNPIN is to provide researchers a standarized platform for training and evaluating algorithms in non-prehensile interactive navigation. BenchNPIN provides simulated environments for a wide range of non-prehensile interactive navigation tasks, metrics specifically capturing both task efficiency and interaction effort, policy templates with reference implementations.

The Environments

Navigating Maze with Movable Obstacles

env = gym.make('maze-NAMO-v0')

This environment features a static maze structure with randomly initialized obstacles. The robot's task is to navigate from a starting position to a goal location while minimizing path length and obstacle collisions.

The Maze environment.

Autonomous Ship Navigation in Icy Waters

env = gym.make('ship-ice-v0')

In this task, an autonomous surface vehicle must reach a horizontal goal line ahead while minimizing collisions with broken ice floes in the channel.

The Ship-Ice environment.

Box Delivery

env = gym.make('box-pushing-v0')

The Box-Delivery environment consists of a set of movable boxes to be delivered to a designated receptacle. The robot is tasked to delivery all boxes using its front bumper.

The Box-Delivery environment.

Area Clearing

env = gym.make('area-clearing-v0')

This envronment consists of a set of movable boxes and a clearance area. The task of the robot is to remove all boxes from this clearance area.

The Area-Clearing environment.

Installation

Install from pip

pip install benchnpin

The pip install above is sufficient to run Ship-Ice and Maze environments. To run Box-Delivery and Area-Clearing, please install shortest path module as follows

git clone https://github.com/IvanIZ/spfa.git
cd spfa
pip install -e .

Build from source

1. Clone the project

git clone https://github.com/IvanIZ/BenchNPIN.git

2. Install dependencies.

cd BenchNPIN
pip install -r requirements.txt

3. Install Gym environment

pip install -e .

4. Install shortest path module

git clone https://github.com/IvanIZ/spfa.git
cd spfa
pip install -e .

Usage

Running an interactive navigation environment

import benchnpin.environments
import gymnasium as gym

env = gym.make('ship-ice-v0')
observation, info = env.reset()

terminated = truncated = False
while not (terminated or truncated):

    action = your_policy(observation)
    observation, reward, terminated, truncated, info = env.step(action)
    env.render()

To configure the parameters for each environment, please refer to the configuration examples for Maze, Ship-Ice, Box-Delivery, and Area-Clearing.

Creating a custom policy from the policy template

from benchnpin.baselines.base_class import BasePolicy

class CustomPolicy(BasePolicy):

    def __init__(self) -> None:
        super().__init__()

        # initialize costum policy here
        ...

    
    def train(self):
        # train the custom policy here, if needed
        ...


    def act(self, observation, **kwargs):
        # define how custom policy acts in the environment
        ...

    
    def evaluate(self, num_eps: int, model_eps: str ='latest'):
        # define how custom policy is evaluated here
        ...

Running benchmarks on policies

from benchnpin.common.metrics.base_metric import BaseMetric
import CustomPolicy1      # some custom policies
import CustomPolicy2
import CustomPolicy3

# initialize policies to be evaluated
policy1 = CustomPolicy1()
policy2 = CustomPolicy2()
policy3 = CustomPolicy3()

# run evaluations
num_eps = 200    # number of episodes to evaluate each policy
benchmark_results = []
benchmark_results.append(policy1.evaluate(num_eps=num_eps))
benchmark_results.append(policy2.evaluate(num_eps=num_eps))
benchmark_results.append(policy3.evaluate(num_eps=num_eps))

# plot efficiency and effort scores
BaseMetric.plot_algs_scores(benchmark_results, save_fig_dir='./')

Implemented Baselines

Tasks	Baselines
Maze	SAC1, PPO1
Ship-Ice	SAC1, PPO1, SAV Planning2, 3
Box-Delivery	SAC1, PPO1, SAM4
Area_Clearing	SAC1, PPO1, SAM4, GTSP5

1: Reinforcement Learning policies Integrated with Stable Baselines 3.

2: Planning-based policy using an ASV ice navigation lattice planner.

3: Planning-based policy using a predictive ASV ice navigation planner.

4: Spatial Action Maps policy.

5: A Generalized Traveling Salesman Problem (GTSP) policy. Please see Appendix I of our paper for details.

Trained Models

You may download the our trained model weights from here.