Hierarchical Reinforcement Learning for Quadrupedal Robots: Efficient Object Manipulation in Constrained Environments
This study introduces a hierarchical reinforcement learning (RL) framework tailored to object manipulation tasks by quadrupedal robots, emphasizing their real-world deployment. The proposed approach adopts a sensor-driven control structure capable of addressing challenges in dense and cluttered envi...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-03-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/25/5/1565 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850052411078475776 |
|---|---|
| author | David Azimi Reza Hoseinnezhad |
| author_facet | David Azimi Reza Hoseinnezhad |
| author_sort | David Azimi |
| collection | DOAJ |
| description | This study introduces a hierarchical reinforcement learning (RL) framework tailored to object manipulation tasks by quadrupedal robots, emphasizing their real-world deployment. The proposed approach adopts a sensor-driven control structure capable of addressing challenges in dense and cluttered environments filled with walls and obstacles. A novel reward function is central to the method, incorporating sensor-based obstacle observations to optimize the decision-making. This design minimizes the computational demands while maintaining adaptability and robust functionality. Simulated trials conducted in NVIDIA Isaac Sim, utilizing ANYbotics quadrupedal robots, demonstrated a high manipulation accuracy, with a mean positioning error of 11 cm across object–target distances of up to 10 m. Furthermore, the RL framework effectively integrates path planning in complex environments, achieving energy-efficient and stable operations. These findings establish the framework as a promising approach for advanced robotics requiring versatility, efficiency, and practical deployability. |
| format | Article |
| id | doaj-art-5e1d067039d542c9be1771d8767a3bb7 |
| institution | DOAJ |
| issn | 1424-8220 |
| language | English |
| publishDate | 2025-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj-art-5e1d067039d542c9be1771d8767a3bb72025-08-20T02:52:49ZengMDPI AGSensors1424-82202025-03-01255156510.3390/s25051565Hierarchical Reinforcement Learning for Quadrupedal Robots: Efficient Object Manipulation in Constrained EnvironmentsDavid Azimi0Reza Hoseinnezhad1School of Information Technology, Deakin University, Victoria 3125, AustraliaSchool of Engineering, RMIT University, Victoria 3082, AustraliaThis study introduces a hierarchical reinforcement learning (RL) framework tailored to object manipulation tasks by quadrupedal robots, emphasizing their real-world deployment. The proposed approach adopts a sensor-driven control structure capable of addressing challenges in dense and cluttered environments filled with walls and obstacles. A novel reward function is central to the method, incorporating sensor-based obstacle observations to optimize the decision-making. This design minimizes the computational demands while maintaining adaptability and robust functionality. Simulated trials conducted in NVIDIA Isaac Sim, utilizing ANYbotics quadrupedal robots, demonstrated a high manipulation accuracy, with a mean positioning error of 11 cm across object–target distances of up to 10 m. Furthermore, the RL framework effectively integrates path planning in complex environments, achieving energy-efficient and stable operations. These findings establish the framework as a promising approach for advanced robotics requiring versatility, efficiency, and practical deployability.https://www.mdpi.com/1424-8220/25/5/1565reinforcement learningrobotic manipulationquadrupedal robots |
| spellingShingle | David Azimi Reza Hoseinnezhad Hierarchical Reinforcement Learning for Quadrupedal Robots: Efficient Object Manipulation in Constrained Environments Sensors reinforcement learning robotic manipulation quadrupedal robots |
| title | Hierarchical Reinforcement Learning for Quadrupedal Robots: Efficient Object Manipulation in Constrained Environments |
| title_full | Hierarchical Reinforcement Learning for Quadrupedal Robots: Efficient Object Manipulation in Constrained Environments |
| title_fullStr | Hierarchical Reinforcement Learning for Quadrupedal Robots: Efficient Object Manipulation in Constrained Environments |
| title_full_unstemmed | Hierarchical Reinforcement Learning for Quadrupedal Robots: Efficient Object Manipulation in Constrained Environments |
| title_short | Hierarchical Reinforcement Learning for Quadrupedal Robots: Efficient Object Manipulation in Constrained Environments |
| title_sort | hierarchical reinforcement learning for quadrupedal robots efficient object manipulation in constrained environments |
| topic | reinforcement learning robotic manipulation quadrupedal robots |
| url | https://www.mdpi.com/1424-8220/25/5/1565 |
| work_keys_str_mv | AT davidazimi hierarchicalreinforcementlearningforquadrupedalrobotsefficientobjectmanipulationinconstrainedenvironments AT rezahoseinnezhad hierarchicalreinforcementlearningforquadrupedalrobotsefficientobjectmanipulationinconstrainedenvironments |