In recent years, robots in diverse forms have been employed to solve various challenging tasks, especially those which have constrained access due to risks. However, before robots can carry out the given task in an environment, they need first to explore and characterize the environment. For an environment with little or no prior information, robots need to use techniques not only limited to camera surveys, sonar scans, LiDAR scans, measuring radiation, etc., to explore and map the surrounding environment.
Moreover, for many circumstances in which the environment is complex and the Global Navigation Satellite System (GNSS) is denied, single robots may be insufficient to fulfill specified tasks. They will need to be controlled in connection with their neighboring team members and utilize neighboring information to localize themselves and generate an overall map. The research outcomes can directly benefit many application scenarios prohibited to human operators; specific scenarios include a facility inspection after a nuclear incident, such as Fukushima Daiichi in 2011, autonomous mining, and seafloor profile plotting.
When a group of networked robots is maneuvering in an unknown space, the proposed control strategy should guarantee that the robots are motivated to explore unvisited areas while avoiding losing connections with other member(s). During the exploration, the robots use various sensors to perceive the environment, localize themselves, and fuse their perceptions through information sharing to build a global map. To this end, many open research questions regarding networked control can be raised. For instance, a collaborative coverage control strategy can be devised to enable robots to explore the space while maintaining their connectivity. Each robot is responsible for watching and sending its local surrounding areas.
Realistic constraints such as limited communication and sensing ranges can be considered when designing the control scheme. Specific control-lite strategies based on learning methods, such as reinforcement learning and deep learning, can be employed to motivate robots to more actively and effectively explore the unvisited areas. Communications among connected robots may also tackle the localization problem. With the input of robot perceptions, the designed control schemes lead to the full exploration of the entire unknown space. This Research Topic aims to present the recent research advances in these areas.
This Research Topic aims to disseminate the latest research achievements in aerial, space, ground, and underwater robotics, with particular emphasis on the multi-robot exploration and mapping in unknown areas. Potential topics include, but are not limited to:
- Design of coordination control strategy of multiple aerial, space, ground, and underwater robots
- Multi-robot simultaneous localization and mapping
- Shared perception in multi-agent collaborative mapping
- Information fusion between connected robots
- Motion planning for exploration
- Reinforcement learning-based multi-agent exploration
- Variable autonomy, shared control, and mixed-initiative systems
- Vision, sensing, perception, navigation, and object detections
- Cooperative and networked robots
- Collaborative robots and human-robot interaction
Keywords:
Multi-robot systems, Networked robots, Autonomous robots, Deep reinforcement learning, Simultaneous localization and mapping, Autonomous exploration, Cyber-physical systems, Shared perception, Coverage control, Distributed control, Robot path planning, Decentralize
Important Note:
All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.
In recent years, robots in diverse forms have been employed to solve various challenging tasks, especially those which have constrained access due to risks. However, before robots can carry out the given task in an environment, they need first to explore and characterize the environment. For an environment with little or no prior information, robots need to use techniques not only limited to camera surveys, sonar scans, LiDAR scans, measuring radiation, etc., to explore and map the surrounding environment.
Moreover, for many circumstances in which the environment is complex and the Global Navigation Satellite System (GNSS) is denied, single robots may be insufficient to fulfill specified tasks. They will need to be controlled in connection with their neighboring team members and utilize neighboring information to localize themselves and generate an overall map. The research outcomes can directly benefit many application scenarios prohibited to human operators; specific scenarios include a facility inspection after a nuclear incident, such as Fukushima Daiichi in 2011, autonomous mining, and seafloor profile plotting.
When a group of networked robots is maneuvering in an unknown space, the proposed control strategy should guarantee that the robots are motivated to explore unvisited areas while avoiding losing connections with other member(s). During the exploration, the robots use various sensors to perceive the environment, localize themselves, and fuse their perceptions through information sharing to build a global map. To this end, many open research questions regarding networked control can be raised. For instance, a collaborative coverage control strategy can be devised to enable robots to explore the space while maintaining their connectivity. Each robot is responsible for watching and sending its local surrounding areas.
Realistic constraints such as limited communication and sensing ranges can be considered when designing the control scheme. Specific control-lite strategies based on learning methods, such as reinforcement learning and deep learning, can be employed to motivate robots to more actively and effectively explore the unvisited areas. Communications among connected robots may also tackle the localization problem. With the input of robot perceptions, the designed control schemes lead to the full exploration of the entire unknown space. This Research Topic aims to present the recent research advances in these areas.
This Research Topic aims to disseminate the latest research achievements in aerial, space, ground, and underwater robotics, with particular emphasis on the multi-robot exploration and mapping in unknown areas. Potential topics include, but are not limited to:
- Design of coordination control strategy of multiple aerial, space, ground, and underwater robots
- Multi-robot simultaneous localization and mapping
- Shared perception in multi-agent collaborative mapping
- Information fusion between connected robots
- Motion planning for exploration
- Reinforcement learning-based multi-agent exploration
- Variable autonomy, shared control, and mixed-initiative systems
- Vision, sensing, perception, navigation, and object detections
- Cooperative and networked robots
- Collaborative robots and human-robot interaction
Keywords:
Multi-robot systems, Networked robots, Autonomous robots, Deep reinforcement learning, Simultaneous localization and mapping, Autonomous exploration, Cyber-physical systems, Shared perception, Coverage control, Distributed control, Robot path planning, Decentralize
Important Note:
All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.