Drones that are capable of autonomous flight in the environment where GPS is not available can be adequately used for various applications such as search and rescue for survivors in the areas hard to reach by conventional methods. This is one of the most active areas of drone research. Prof. David Hyunchul Shim’s team has won the First prize in the AI Grand Challenge hosted by MOSCIT. His research group has been collaborating with NASA JPL as a member of CoSTAR, which won 1st place in last Phase 2 competition.


Participation in AI Grand Challenge hosted by MOSICT

Prof. Shim’s research group has been working on the autonomous indoor flight technology using LiDAR-based SLAM in complex indoor environment. Such efforts enabled his team to win the AI Grand Challenge host by MOSICT in July 2019, where a drone is required to fly in an arena, which is filled with poles, a tunnel, trees, nets, a wind blower, and more. The morale behind this competition is to develop a number of AI-powered technologies to eventually construct a system to automatically find survivors and then eventually find and rescue them using a drone, developed by the winning team. The team was able to win the competition using their own recipe of a drone system that is purpose-built for the indoor operation. The prize of this competition is 1 Million USD for additional research and the Minister’s award from MOSICT. Also, Prof. Shim’s research group has collaborated with NASA JPL to participate in DARPA Subterranean Challenge, which requires a group of robots, both ground and aerial to autonomously navigate in complex subterranean environments such as tunnels, factories, and natural caves. Their collaboration helped Prof. Shim’s group to broaden their view and management know-hows on the subterranean operations.


Future outlook of indoor flight

Autonomous indoor flight is a very active research area and will play an important role for real-world applications. Such a technology will enable the drone to perform its mission such as search and rescue using its own decision in a complex environment. Especially, such drones will be used for complex disaster, underground exploration, maritime rescues, where human rescuers may have great difficulty. Their research results are highly expected to be used in such applications with a great success.


Prof. Shim, Hyunchul
2019 KI Annual Report