Reference : Towards incremental autonomy framework for on-orbit vision-based grasping |
Scientific congresses, symposiums and conference proceedings : Paper published in a book | |||
Engineering, computing & technology : Aerospace & aeronautics engineering | |||
http://hdl.handle.net/10993/49355 | |||
Towards incremental autonomy framework for on-orbit vision-based grasping | |
English | |
Barad, Kuldeep Rambhai ![]() | |
Martinez Luna, Carol [University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > Space Robotics >] | |
Dentler, Jan [Redwire Space Europe > Robotics] | |
Olivares Mendez, Miguel Angel [University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > Space Robotics >] | |
29-Oct-2021 | |
Proceedings of the International Astronautical Congress, IAC-2021 | |
No | |
No | |
International | |
International Astronautical Congress | |
24-10-2021 to 29-10-2021 | |
International Astronautical Federation | |
Dubai | |
United Arab Emirates | |
[en] OSAM ; Autonomy, On-orbit Robotics, Perception ; , Manipulation, Learning | |
[en] This work presents a software-oriented autonomy framework that enables the incremental development of high
robotic autonomy. The autonomy infrastructure in space applications is often cost-driven and built for a narrow time/complexity domain. In domains like On-orbit Servicing Assembly and Manufacturing (OSAM), this prevents scalability and generalizability, motivating a more consistent approach for the incremental development of robotic autonomy. For this purpose, the problem of vision-based grasping is described as a building block for high autonomy of dexterous space robots. Subsequently, the need for a framework is highlighted to enable bottom-up development of general autonomy with vision-based grasping as the starting point. The preliminary framework presented here comprises three components. First, an autonomy level classification provides a clear description of the autonomous behavior of the system. The stack abstraction provides a general classification of the development layers. Finally, the generic execution architecture condenses the flow of translating a high-level task description into real-world sense-planact routines. Overall, this work lays down foundational elements towards development of general robotic autonomy for scalablity in space application domains like OSAM. | |
http://hdl.handle.net/10993/49355 | |
FnR ; FNR15799985 > Kuldeep Rambhai Barad > MIS-URSO > Modular Vision For Dynamic Grasping Of Unknown Resident Space Objects > 01/04/2021 > 15/01/2025 > 2021 |
File(s) associated to this reference | ||||||||||||||
Fulltext file(s):
| ||||||||||||||
All documents in ORBilu are protected by a user license.