KI for Robotics

  • home_icon HOME
  • Research
  • Research Institutes
  • KI for Robotics
  • Highlight

Total 12

Study on Cooperative Autonomous Multi-UAV Systems

Dawn of the era of unmanned autonomous systems

“Needs and applications for unmanned aerial vehicles (UAVs) are growing rapidly beyond all expectations. This study successfully developed a technology that can be directly applied to the operation of unmanned aerial vehicles for defense purposes. It appears that cooperation between manned and unmanned aerial vehicles in missions as well as functional networks of various numbers and multiple layers of unmanned aerial vehicles for surveillance and reconnaissance will become possible in the near future. The topic that most interests me at present is the combination of 'safety' and 'security' in the cooperation of unmanned aerial vehicles. It is of the utmost importance to ensure safety in order to implement unmanned autonomous technologies that benefit humankind. I plan to continue research towards that goal.”

Development of an algorithm for cooperative missions between multiple unmanned aerial vehicles

Aerospace research to collect information over a wide geographical area for tracking missions has been conducted for many years. Recently, interest has converged on the utilization of multiple unmanned aerial vehicles/robots to conduct cooperative missions. For manned and unmanned aerial vehicles with sensors to collect information, they have to be connected in a single network and operated with an algorithm. Professor Choi Han-Lim's team at KIDCS has developed a core algorithm for this purpose to construct a cooperative system for heterogeneous multi-agents, signaling the beginning of an era of unmanned autonomy to enhance the quality of human lives.

The project entitled “Preliminary study on cooperative systems of heterogeneous autonomous multiple UAVs” was funded by the Agency for Defense Development and carried out as a joint project of seven professors from KIDCS and the Department of Aerospace Engineering between Professor Choi's team. The scope of the project included development of fundamental theories and algorithms for cooperation of multiple UAVs with heterogeneous characteristics/capabilities, and also hardware-based technology demonstration for validation of the theories and algorithms. The role of each participating research team was defined based on their expertise to align with the common goal to make innovation on cooperative multi-UAV systems.

The core of this study lies in the realization of a high level of autonomy. Unmanned aerial vehicles in current use are operated remotely by pilots on the ground through information obtained from sensors. However, in the near future, autonomy could be developed to the human supervision level, where human decision-making would be limited to defining the mission and unmanned drones would perform the actual mission autonomously based on their own analyses of the field situation. The system would be able to perform various missions under human approval and direction at the highest level while both humans and the system would be able to reset the system based on observational data.

Professor Choi Han-Lim and his team focused on three questions: firstly, the selection of problems and methods for problem formulation; secondly, which algorithm to use to achieve autonomous decision-making in a dispersed environment; and thirdly, how to effectively apply and verify such technologies in actual aerial vehicles. The earlier part of the study focused on the expansion and generalization of the theoretical framework, while the latter part of the project focused on improving the algorithm based on this expanded theory and applying it to actual hardware to enhance reliability. In the final year, the team contemplated the question of the effective demonstration and verification of the efficacy of the algorithm.

The research process was as follows. An algorithm was developed to plan and assign tasks and missions as well as update plans in real time according to changes in the situation within an environment of multiple dispersed unmanned aerial vehicles. To successfully demonstrate the effectiveness of the technology in complex cooperative mission scenarios, this algorithm was applied to a team of miniature quadrotors and ground robots in an indoor environment with motion capture cameras.

Professor Choi Han-Lim's team is currently focused on the incorporation of safety and security into the field of cooperative unmanned aerial vehicle technology. A world where everything is connected through the Internet of Things offers conveniences but also presents the threat of cyberattacks moving beyond cyberspace and into physical environments. Unmanned aerial vehicles flying overhead or unmanned road vehicles driving past our homes could be hacked into.

Thus, security is the most crucial factor in realizing unmanned autonomous technologies that benefit humankind. If research and development up till now were focused on actualizing unmanned and autonomous technologies themselves, now is the time for further consideration of the application of these technologies in the real world. Professor Choi Han-Lim and his team are working under this overall vision to develop technologies for autonomous decision-making in unmanned aerial vehicles that consider safety, security, and human interaction.

Prof. Choi, Han-Lim
2014 Annual Report

KAIST 291 Daehak-ro, Yuseong-gu, Daejeon (34141)
T : +82-42-350-2381~2384
F : +82-42-350-2080
Copyright (C) 2015. KAIST Institute