Leveraging Expert Demonstration Features for Deep Reinforcement Learning in Floor Cleaning Robot Navigation
In this paper, a Deep Reinforcement Learning (DRL)-based approach for learning mobile cleaning robot navigation commands that leverage experience from expert demonstrations is presented. First, expert demonstrations of robot motion trajectories in simulation in the cleaning robot domain are collecte...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-10-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/22/20/7750 |
| _version_ | 1827648040269250560 |
|---|---|
| author | Reinis Cimurs Emmanuel Alejandro Merchán-Cruz |
| author_facet | Reinis Cimurs Emmanuel Alejandro Merchán-Cruz |
| author_sort | Reinis Cimurs |
| collection | DOAJ |
| description | In this paper, a Deep Reinforcement Learning (DRL)-based approach for learning mobile cleaning robot navigation commands that leverage experience from expert demonstrations is presented. First, expert demonstrations of robot motion trajectories in simulation in the cleaning robot domain are collected. The relevant motion features with regard to the distance to obstacles and the heading difference towards the navigation goal are extracted. Each feature weight is optimized with respect to the collected data, and the obtained values are assumed as representing the optimal motion of the expert navigation. A reward function is created based on the feature values to train a policy with semi-supervised DRL, where an immediate reward is calculated based on the closeness to the expert navigation. The presented results show the viability of this approach with regard to robot navigation as well as the reduced training time. |
| first_indexed | 2024-03-09T19:30:49Z |
| format | Article |
| id | doaj.art-8ac990aa1b8947f6b1c1e58a0a85144f |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-09T19:30:49Z |
| publishDate | 2022-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-8ac990aa1b8947f6b1c1e58a0a85144f2023-11-24T02:25:10ZengMDPI AGSensors1424-82202022-10-012220775010.3390/s22207750Leveraging Expert Demonstration Features for Deep Reinforcement Learning in Floor Cleaning Robot NavigationReinis Cimurs0Emmanuel Alejandro Merchán-Cruz1SIA Robotic Solutions, LV-1039 Riga, LatviaSIA Robotic Solutions, LV-1039 Riga, LatviaIn this paper, a Deep Reinforcement Learning (DRL)-based approach for learning mobile cleaning robot navigation commands that leverage experience from expert demonstrations is presented. First, expert demonstrations of robot motion trajectories in simulation in the cleaning robot domain are collected. The relevant motion features with regard to the distance to obstacles and the heading difference towards the navigation goal are extracted. Each feature weight is optimized with respect to the collected data, and the obtained values are assumed as representing the optimal motion of the expert navigation. A reward function is created based on the feature values to train a policy with semi-supervised DRL, where an immediate reward is calculated based on the closeness to the expert navigation. The presented results show the viability of this approach with regard to robot navigation as well as the reduced training time.https://www.mdpi.com/1424-8220/22/20/7750Deep Reinforcement Learningmobile robot navigationsemi-supervised learningautonomous cleaning robots |
| spellingShingle | Reinis Cimurs Emmanuel Alejandro Merchán-Cruz Leveraging Expert Demonstration Features for Deep Reinforcement Learning in Floor Cleaning Robot Navigation Sensors Deep Reinforcement Learning mobile robot navigation semi-supervised learning autonomous cleaning robots |
| title | Leveraging Expert Demonstration Features for Deep Reinforcement Learning in Floor Cleaning Robot Navigation |
| title_full | Leveraging Expert Demonstration Features for Deep Reinforcement Learning in Floor Cleaning Robot Navigation |
| title_fullStr | Leveraging Expert Demonstration Features for Deep Reinforcement Learning in Floor Cleaning Robot Navigation |
| title_full_unstemmed | Leveraging Expert Demonstration Features for Deep Reinforcement Learning in Floor Cleaning Robot Navigation |
| title_short | Leveraging Expert Demonstration Features for Deep Reinforcement Learning in Floor Cleaning Robot Navigation |
| title_sort | leveraging expert demonstration features for deep reinforcement learning in floor cleaning robot navigation |
| topic | Deep Reinforcement Learning mobile robot navigation semi-supervised learning autonomous cleaning robots |
| url | https://www.mdpi.com/1424-8220/22/20/7750 |
| work_keys_str_mv | AT reiniscimurs leveragingexpertdemonstrationfeaturesfordeepreinforcementlearninginfloorcleaningrobotnavigation AT emmanuelalejandromerchancruz leveragingexpertdemonstrationfeaturesfordeepreinforcementlearninginfloorcleaningrobotnavigation |