A Multi-Resolution Frontier-Based Planner for Autonomous 3D Exploration

Batinović, Ana; Petrović, Tamara; Ivanović, Antun; Petric, Frano; Bogdan, Stjepan

doi:10.1109/lra.2021.3068923

Cited by 49 publications

(40 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [1], authors proposed the NBV planner (RH-NBVP), which uses an RRT-based search [3], [4] to guide a UAV into the unexplored area. While the method showed good scaling properties and performance in a local exploration, it is not resilient to dead ends, resulting in a poor global scene coverage and thus, a high overall exploration time, as shown in [13], [22], [17], and in our previous work [16]. To address the drawbacks of the RH-NBVP, Witting et al [9] introduced several modifications: memorizing previously visited locations; local gain optimization; and trajectory optimization, resulting in faster exploration.…”

Section: Related Workmentioning

confidence: 85%

See 1 more Smart Citation

A Shadowcasting-Based Next-Best-View Planner for Autonomous 3D Exploration

Batinović¹,

Ivanović²,

Petrović³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

In this paper, we address the problem of autonomous exploration of unknown environments with an aerial robot equipped with a sensory set that produces large point clouds, such as LiDARs. The main goal is to gradually explore an area while planning paths and calculating information gain in short computation time, suitable for implementation on an on-board computer. To this end, we present a planner that randomly samples viewpoints in the environment map. It relies on a novel and efficient gain calculation based on the Recursive Shadowcasting algorithm. To determine the Next-Best-View (NBV), our planner uses a cuboid-based evaluation method that results in an enviably short computation time. To reduce the overall exploration time, we also use a dead end resolving strategy that allows us to quickly recover from dead ends in a challenging environment. Comparative experiments in simulation have shown that our approach outperforms the current state-of-the-art in terms of computational efficiency and total exploration time. The video of our approach can be found at https://www.youtube.com/playlist?list=PLC0C6uwoEQ8ZDhny1V dmFXLeTQOSBibQl.

show abstract

Section: Related Workmentioning

confidence: 85%

“…It is shown that this approach outperforms the closest frontier method [11]. Frontier-based exploration approaches for 3D environments are also researched in [14], [15], [16], [17], [18], [19].…”

Section: Related Workmentioning

confidence: 99%

A Shadowcasting-Based Next-Best-View Planner for Autonomous 3D Exploration

Batinović¹,

Ivanović²,

Petrović³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Generating waypoints for navigation planners is an essential aspect of robot team autonomy. In our work, several methods were implemented: (a) waypoint generation using a 2D lawnmover pattern in relatively small areas of known size; (b) Levvy flight 2D waypoint generation for large areas of known size [19]; (c) autonomous 2D [20] and 3D exploration [21] for areas of unknown size. In this work, we briefly discuss our autonomous exploration approach for ERL and MBZIRC competitions, which was later extended to the planner described in [21], based on the exploration tool called 3D-FBET, work of Zhu et al [22].…”

Section: Related Workmentioning

confidence: 99%

From ERL to MBZIRC: Development of An Aerial-Ground Robotic Team for Search and Rescue

Arbanas¹,

Petric²,

Batinović³

et al. 2022

Automation and Control - Theories and Applications

Self Cite

View full text Add to dashboard Cite

This chapter describes the efforts of the LARICS team in the 2019 European Robotics League (ERL) Emergency Robots and the 2020 Mohamed Bin Zayed International Robotics Challenge (MBZIRC) robotics competitions. We focus on the implementation of hardware and software modules that enable the deployment of aerial-ground robotic teams in unstructured environments for joint missions. In addition to the overall system specification, we outline the main algorithms for operation in such conditions: autonomous exploration of unknown environments and detection of objects of interest. Analysis of the results shows the success of the developed system in the competition arena of two of the largest outdoor robotics challenges. Throughout the chapter, we highlight the evolution of the robotic system based on the experience gained in the ERL competition. We conclude the chapter with key findings and additional improvement ideas to advance the state of the art in search and rescue applications of heterogeneous robotic teams.

show abstract

“…The well-established family of approaches incorporates the concept of next best pose process, i.e., a turn-based, greedy selection of the next best position (also known as frontiercell) to acquire measurement, based on heuristic strategy (e.g., [5][6][7]). Although this family of approaches has been extensively studied, some inherent drawbacks significantly constrain its broader applicability.…”

Section: Related Workmentioning

confidence: 99%

“…Partial Observability. Due to the nature of the exploration/coverage setup, at each timestep, the robot is only aware of the location of the obstacles that have been sensed from the beginning of the episode (5). Therefore, any long-term plan should be agile enough to be adjusted on the fly, based on future information about the unknown obstacles' positions.…”

Section: Key Rl Attributesmentioning

confidence: 99%

MarsExplorer: Exploration of Unknown Terrains via Deep Reinforcement Learning and Procedurally Generated Environments

et al. 2021

View full text Add to dashboard Cite

This paper is an initial endeavor to bridge the gap between powerful Deep Reinforcement Learning methodologies and the problem of exploration/coverage of unknown terrains. Within this scope, MarsExplorer, an openai-gym compatible environment tailored to exploration/coverage of unknown areas, is presented. MarsExplorer translates the original robotics problem into a Reinforcement Learning setup that various off-the-shelf algorithms can tackle. Any learned policy can be straightforwardly applied to a robotic platform without an elaborate simulation model of the robot’s dynamics to apply a different learning/adaptation phase. One of its core features is the controllable multi-dimensional procedural generation of terrains, which is the key for producing policies with strong generalization capabilities. Four different state-of-the-art RL algorithms (A3C, PPO, Rainbow, and SAC) are trained on the MarsExplorer environment, and a proper evaluation of their results compared to the average human-level performance is reported. In the follow-up experimental analysis, the effect of the multi-dimensional difficulty setting on the learning capabilities of the best-performing algorithm (PPO) is analyzed. A milestone result is the generation of an exploration policy that follows the Hilbert curve without providing this information to the environment or rewarding directly or indirectly Hilbert-curve-like trajectories. The experimental analysis is concluded by evaluating PPO learned policy algorithm side-by-side with frontier-based exploration strategies. A study on the performance curves revealed that PPO-based policy was capable of performing adaptive-to-the-unknown-terrain sweeping without leaving expensive-to-revisit areas uncovered, underlying the capability of RL-based methodologies to tackle exploration tasks efficiently.

show abstract

A Multi-Resolution Frontier-Based Planner for Autonomous 3D Exploration

Cited by 49 publications

References 23 publications

A Shadowcasting-Based Next-Best-View Planner for Autonomous 3D Exploration

A Shadowcasting-Based Next-Best-View Planner for Autonomous 3D Exploration

From ERL to MBZIRC: Development of An Aerial-Ground Robotic Team for Search and Rescue

MarsExplorer: Exploration of Unknown Terrains via Deep Reinforcement Learning and Procedurally Generated Environments

Contact Info

Product

Resources

About