GIST FOCUS

GIST opens a new era of autonomous vehicle!

In preparation for the 4th Industrial Revolution, GIST's activities, which are creating an AI-based start-up campus, have recently caught fire. The fastest moving place are the faculties' labs that are conducting research on artificial intelligence. Currently,
GIST professors are focusing on developing AI convergence and applied technologies. We visited the Machine Learning and
Vision Lab of Professor Moongu Jeon in the School of Electrical Engineering and Computer Science to understand its
autonomous driving project that it is preparing to demonstrate in December.
Machine Learning and Vision Lab

Professor Moongu Jeon
School of Electrical Engineering and Computer
All set for autonomous driving
Professor Moongu Jeon, who is in charge of the Machine Learning and Vision Lab in the School of Electrical Engineering and
Computer, has mainly studied technology such as machine learning, computer vision, pattern recognition, etc. and has achieved various research achievements with visual surveillance technologies, especially in the security field.
Professor Moongu Jeon elaborated by saying, "Over the past few years, I have been working with vision technology to detect
people's faces and movements using CCTV cameras, as well as capturing illegal or abnormal behavior. When these vision
technologies, which are used in the security field, are combined with automotive technologies, they can detect and predict the movements of vehicles and people, which is highly related to autonomous driving.”
In other words, autonomous driving technology requires artificial intelligence, which is a research area of Professor Jeon,
who has already acquired the basic technologies necessary for autonomous driving. Professor Jeon said that if autonomous
driving technology is successfully realized, it will not only solve the current traffic problems but also save many lives lost in
traffic
Autonomous driving demonstration planned at the end of this year in December
There are many different technology required for successful autonomous driving, which includes camera vision, LiDAR, GPS, and
machine learning as well as sensor technology for identifying possible obstacles, algorithm for comprehensively analyzing data to find an optimized route to a destination, and deep learning for handling unexpected situations on the way to the
destination.

To continuously expand and improve his research, Professor Moongu has been cooperating with School of Electrical Engineering and Computer Science Professor Jong Won Shin and Professor Chang-Soo Park as well as School of Mechanical Engineering Professor Hyo Sung Ahn
Professor Jong Won Shin of the Multimedia Signal Processing Lab is responsible for the audio GUI technology applied to autonomous driving vehicles. Professor Chang-Soo Park from the Wired and Wireless Communication Systems Lab is developing the new LiDAR sensor technology. Professor Hyo Sung Ahn is working on the platooning driving technology to efficiently implement multiple autonomous driving vehicles at once. Through this cooperation, the objective is to introduce a car that can drive autonomously, if not fully self-driving, by December of this year, with various technologies having been completed such as teleoperation, localization, and other basic GUI
Professor Moongu Jeon stated that an autonomous driving shuttle project is also planned inside GIST once the autonomous driving demonstration is successfully completed this year and its safety can be guaranteed. In addition, if the objectives set for this year is accomplished, further development will be made also regarding technology to minimize the traffic accidents advancing the existing technology to the next stage and enabling information exchange among autonomous vehicles.
Professor Jeon said, "I plan to form and lead an autonomous driving contest club for undergraduates and graduates in the near future. I hope there will be heartful support from GIST as well as interests from students for this autonomous driving project."
“To make the autonomous driving demonstration in December, various AI convergence technologies are under research and development. Once this is successfully completed, we plan to create an autonomous shuttle project inside GIST.”
Keyword Interview
  Keyword1  
Autonomous driving technology
Besides AI and vision technology, there are multiple technologies required for autonomous driving, such as fusing various sensor data and integrating the overall system. We plan to show you the autonomous driving vehicle within this year, and we are well on our way to develop the following sets of technology.
  1     Detecting/recognizing a small object (obstacle) from the camera and LiDAR data
  2     Recognizing surrounding vehicles by analyzing data from various sensors
  3     Recognizing the current location and position to finding the optimal route to a destination
  4     Collecting vehicle data
  5     GUI display data for current vehicle condition
  6     Connecting sensors and computers with the vehicle
  Keyword2  
Challenges
One of the toughest challenges for this project was that there is no specific information on the vehicle itself. We had to collect the required information on the internal data of the vehicle to implementing autonomous driving. Data however was difficult to comprehend, unless you are the original manufacturer. We found it very difficult to figure out from the information on how to use a computer to control the brake, the accelerator, and the steering of the electric vehicl. As a solution, we could not help but purchase an interface device manufactured by a third party that was capable of handling these issues.
On the other hand, the preliminary challenge was resolving economic and legal issues before actually using being able to use an autonomous vehicle. Regarding the economic sideissues, it is very expensive to purchase the many sensors needed for an autonomous driving so it is necessary to lower these costs before autonomous vehicles can become economically practical.
Also, taking the recent Uber autonomous driving accident as a precedent, there should be legal guidelines to determine on who is at fault and legally responsible once there is a human casualty out of thea person is hurt in an autonomous driving accident. Unless these two preliminaries challenges are solved, it will take more time to realize widespread autonomous driving even if technological issues are overcome.
  Keyword3  
Benefits
Once autonomous vehicles are successfully implemented, we expect that the road traffic will be improved while many lives will be saved from accidents. Additionally, since driver's licenses will no longer be needed in addition, the elderly will benefit from being able to drive by themselves.
The residential environment will also be affected in the future because there will be no need to live in an area that is easily accessible to public transportation, so people will start moving to suburban locations with better living standards.

GIST's 25th anniversary synergy keyword is 'AI'
AI ecosystem for research, education, and job creation

Korea's first AI-based job creation complex is expected to be established in Gwangju with the purpose of developing AI
technology, the core strength of the 4th Industrial Revolution.
GIST has been chosen as the advance planning organization for this job creation complex that will be located in the Gwangju
Cheomdan 3 District. Being at the center of this project, GIST will lead the creation of new values derived from AI with excellent
research results and educational innovation.
The AI industry is at the center of future growth engines
Amid the rapid increase in global investment and demand for technology development regarding AI and its convergence with
existing industries, the establishment of a start-up complex for fostering the AI industry, which will become a future growth
engine, has been required.
Major developed countries are already moving rapidly and are making significant investments around existing AI industry
ecosystems. The United States will establish a professional research institute with the help of the government and academia to
actively develop AI technology with the support of $3 billion for 10 years starting in 2013 through the Brain Initiative Plan.
Japan established the AI Research Center to create a seamless cycle for AI development, commercialization, and convergence
research with support of 100 billion yen for 10 years starting in 2016. Korea plans to an AI-based start-up complex in Gwangju,
which is one of the 5-year government administration plans, and is in its final stages of implementation for 2019.
  point  1   GIST creates AI-based start-up complex
This project is to form a 660,000 cubic meterthe AI-based start-up complex with in the Cheomdan 3 District next to GIST for
10 years from 2019 to 2028. Along with GIST, different industrial and educational institutes in Gwangju have joined including
Gwangju City, Gwangju Information & Culture Industry Promotion Agency, Korea Testing Laboratory, Gwangju Technopark, etc.
GIST, which is at the center of it, has been laying the groundwork for carrying out plans to create an AI-based start-up complex
over the past few years by increasing its AI-related research capabilities. It currently operates the GIST Artificial Intelligence
Center, SW Education Center and Intelligent Robot Center, along with 52 artificial intelligence-related professors and
researchers. In addition, GIST has actively participated in global AI research, including publishing 36 AI papers, carrying out 7.2
billion won worth of research tasks, and gaining 62 patents related to AI over the past five years, allowing GIST to have a
pivotal role in the creation of AI start-up complex. The ‘Maeil Business Newspaper-GIST forum’ was also held on July 11 at
GIST's Oryong Hall and various discussions were held about GIST's role in the development of AI technology under the theme
of 'Future Social Changes Caused by AI and GIST's Response.'
  point  2   Synergy among research, education, and job creation for AI industry development
The ‘AI-based start-up complex is expected to have investment of total KRW 1 trillion and will have 3 major tasks: establishing
an AI research institute, creating an AI campus, and developing an AI industrial ecosystem. This is part of a bigger plan to
develop the AI industry through synergy among research, education, and job creation while harmonizing 3 different aspects by
establishing a domestic/foreign industrial and educational research cooperation system, creating an educational system
for promoting convergence personnel required for the 4th Industrial Revolution, and strengthening the capability of new AI
companies.
The AI research institute will be established as a large convergence research institute that differentiates itself from existing
universities and institutions with world-class artificial intelligence-based technology and will conduct research on core
technologies for corporate projects and the 4th Industrial Revolution.
For the AI campus, an AI specialized human resources curriculum will be implemented to train the convergence talent needed
for the 4th Industrial Revolution while the creation of an AI start-up ecosystem will serve as the basis for strengthening
corporate capabilities by building a start-up complex and enhancing global competitiveness
Another accomplishment expected for this project is that by creating a successful model of this AI-based start-up complex
in Korea, it would become a hub connected to other AI clusters in different regions. Starting with the AI-based start-up complex
in Gwangju, a national AI ecosystem can be actualized with specific cooperation among research, education, and job creation
promotion institutes in other regions such as Chungcheong and Gangwon. In addition, self-reliance measures for the
development of a sustainable AI start-up complex will be developed after the completion of the project to nurture AI
convergence talent based on world-class AI R&D technology and to promote trained talents in the global market through
the creation of AI technologies. Start-up companies with global competitiveness are expected to develop AI R&D with global
competitiveness, creating a virtuous AI-based start-up complex ecosystem.
  point  3   The goal of creating the world's best AI-based start-up complex
There is a worldwide trend at the moment ato build strong industrial ecosystems that are focused on universities capable of
conducting researches in specific fields. If GIST's AI research and development capabilities are linked to major industries and
infrastructures in the region, it will be possible to find and foster future industries based on AI.
This project to create an AI R&D hub at GIST by establishing the AI-based start-up complex, AI campus, AI research institute,
and job creation center has the goal to make the world’s best AI-based start-up complex ecosystem, nurturing the core
personnel who can lead the R&D for the 4th Industrial Revolution and can lead the creation of business ecosystem.
The AI-based start-up complex is mainly located in Gwangju and Jeonnam, and it is expected to contribute by creating future
jobs through AI technology development, which is the core of 4th Industrial Revolution in Korea. We look forward to seeing the
city of Gwangju, which is making big plans to develop into a 4th Industrial Revolution convergence center, and GIST leading the development of the world's best AI technology at the center of this project.