● 인공지능-로봇 융합연구분야
인공지능-로봇 융합 연구 분야는 고도화된 인공지능 및 로봇 기술을 원자력시설 안전검사, 원자력시 설 자동화, 원자력시스템 무인화 등에 적용할 수 있는 융합기술 개발을 목표로 연구를 수행하고 있다.
원자력분야 응용 목적의 로봇은 일반적인 로봇과 다르게 극한의 환경에서도 안정적으로 동작할 수 있 어야 하기 때문에, 고도화된 하드웨어 제작기술과 강인한 소프트웨어 알고리즘이 필요하다.
하드웨어 제작기술은 수중, 수상, 지상, 공중의 소형 로봇부터 대형 로봇, 특수 목적의 센서 플랫폼 설계 기술을 연구하고 있다.
소프트웨어 기술은 카메라, 초음파, 화학, 적외선카메라 등의 각종 센서의 데이터 처리 기법, 머신러닝 및 딥러닝 기법을 집중적으로 연구하고 있다. 또한 극한환경에서의 로봇의 자율주행 기술 및 제어 기술을 개발하고 있다.
● 원자력-인공지능 융합 연구 분야
Automatic Design Algorithm of Reactor Core
Designing reactor cores by means of an artificial neural network is a difficult challenge, because there are many variables in the core configuration.
This study presents a feasibility study on the automatic design of a research reactor core using an artificial neural network.
By imitating conventional design procedure, a way to design the core is developed by means of the artificial neural network and automatic machine learning.
The results reveal that the reactor core designed by the proposed method performs well and will, therefore, provide a clue to innovation in future reactor design with artificial intelligence.
● 원자력-인공지능 융합 연구 분야
AI Drone System for Fire Detection and Coping
For automatic detection and suppression of fire, AI drone system is under development.
Flaming regions in images obtained by cameras are localized by the embedded system with the segmentation technique.
After analyzing fire area, the guided flight system of drone is activated to the area, and fire extinguisher is automatically dropped.
● 원자력-인공지능 융합 연구 분야
Chromosome Detection System Using Deep-Learning Technique for Radiation Exposure
Dicentric Chromosome Assay (DCA) has been used for estimating the absorbed radiation dose following occupational or incidental radiation exposure.
The estimating procedure of DCA is labor intensive and time consuming, and therefore, it cannot be widely utilized for radiation mass casualty incidents.
Automatic estimation systems of DCA has been studied to improve the efficiency of the estimation procedure in previous studies. These methods, however, cause significant problems on the accuracy of the chromosome detection.
For improving the accuracy and efficiency of the DCA, an automatic estimation system of DCA using deep learning technique is under development in our research group.
● 원자력-인공지능 융합 연구 분야
Autonomous Mobile Drone
Generally, radioactive disposal facilities has low accessibility because of the problem related to the radiation exposure and complexity of the facility.
One of the solution for inspecting the facility is a use of drone; however, automatic drone controlling without GPS as well as considering the various obstacles is a difficult challenge.
Our team is trying to develop an autonomous control system of the drone with SLAM, ROS and deep learning technique.
● 인공지능-로봇 융합연구분야
Nuclear Power Plant’s Dome Inspection Robot
This robot system is a cooperative robot system for dome inspection in nuclear power plants (NPP).
It has the ability to perform a grid laser–based ceiling mapping and localization.
Non-destructive inspection is available by a wall-climbing agent robot, which is attached to the surface of the dome utilizing an aerodynamic force. Lifting drone equipped with the grabber mechanism lifts up the wall-climbing agent to the ceiling.
● 인공지능-로봇 융합연구분야
Autonomous Underwater Vehicle ‘Cyclops’
Underwater is one of the hazardous environments, and GPS and RF do not work. ‘Cyclops’ is an underwater robot of the hovering type, which enables precise 3-D position control in the water by using multiple propellers. I
t is equipped with various measurement sensors and navigation sensors such as high-resolution still camera, lighting system, acoustic camera, laser, and chemical sensor, so it can respond to various missions.
It can be used for underwater precision tasks such as underwater environment and ecological investigation, safety inspection of underwater structures.
● 인공지능-로봇 융합연구분야
Robust Perception Using Artificial Intelligence
Robots in hazardous robots use optical or acoustic sensors to sense the surrounding environment. In hazardous environment, there is no light source such as the sun and light is severely attenuated and scattered by fog, dust, and water. Therefore, optical sensors have limited views.
On the other hand, acoustic sensors are independent of visibility but have a limitation of low resolution and noise-to-signal ratio.
To recognize the surrounding environment robustly and perform precise mission in harsh condition, various artificial intelligence-based perception algorithms such as target object recognition, image enhancement, and opti-acoustic fusion are required.
● 인공지능-로봇 융합연구분야
3-D Mapping and Localization with limited visibility
In extreme environments with limited visibility, optical sensors have limited performance. The acoustic sensors can be one of the alternatives to the optical sensors in a low-visibility environment. However, the acoustic sensors have poor SNR characteristics, low-resolution problems, and loss of height information.
Therefore, a proper signal processing technique is required. This research is a 3-D seafloor scanning method using multibeam sonar.
It provides a unique analysis of sonar image geometry for extracting missing elevation information and can be continuously executed regardless of the existence of features.
