Construction IT / Urban / Transportation

Home > RESEARCH > Research Area > Construction IT / Urban / Transportation


Intelligent Robotic Autonomy and Perception Lab (IRAP)
 
SLAM based Urban Mapping

Our group works on 3D urban mapping using cameras, LiDARs, and other inertial sensors. We use two different sensor systems for mapping, depending on the LiDAR configuration, push broom style and 360 tilted scanning style.

1) 3D Map of Urban Canyon Urban 3D map accuracy and consistency heavily rely on the localization accuracy of the mapping vehicle. The vehicle localization is challenging in an urban area where high rise building causes sporadic GPS reception. Even if the GPS is received, the consumer level GPS can merely guarantee meter-level accuracy. An initial and fast solution we suggest is to leverage the existing 3D map and turn them into a dense 3D map. For this research, our focus is at automating the mapping procedure based upon Simultaneous Localization and Mapping (SLAM) technique.

2) Lane map for autonomous vehicles Generation of accurate lane map is a challenge in terms of the vehicle localization. Using a monocular camera for lane map generation is achieved. Furthermore, we present using the generated lane maps for vehicle localization.

 
 
Hull Inspection using Underwater Robot
Visual in-water hull inspection using robot requires various underwater robotics technologies. Since the underwater is a typical GPS denied environment, perceptual or acoustic sensors provide mean to bound the navigational error. Our group works on vision based approach toward the underwater visual SLAM. By visually recognizing the previously visited location, a robot can correct navigation drift and accurately localize within the water medium.
 
 
Underwater navigation using imaging sonar
Although visual sensors provide intuitive and rich information around the world, limited sight often prohibits effective implementation. Sonar in that sense is an effective solution for underwater perception as it provides reliable sensor data regardless of water turbidity. Among many sonar sensors, our group focuses on imaging sonar and apply conventional vision approaches over the sonar images.
 
 
Real-time channel invariant dehazing
We provide an effective software module that is applicable for dehazing. Conventional approaches for dehazing usually exploit color information to estimated required parameters. Our solution is fully real-time (>10Hz) and works equally well for both color and grayscale images. Our method is applicable to various degradation, fog, smoke, and turbid underwater.
 
 
Maturepolis Lab
 
Smart waste systems
Solid waste management is a crucial service in cities, with many implications for public health and sustainability. However, many cities lack useful data on where trash and recyclables are generated, collected, and processed. Our research focuses on ways to apply sensors and data analysis to make waste management more transparent and efficient. We also look at informal systems, particularly how elderly recyclers navigate and collect waste even in well-developed urban areas.
 
 
Computer vision and urban environments

This research uses computer vision (CV) to understand human behavior in public spaces, and iteratively improve the design of those spaces. Building on theories from past urban planners and designers, we apply CV to automating analysis of pedestrians in these spaces. Such data can help us improve the safety, activity, and value of public spaces, and allow us to experiment with improvements to design. We also look at ways to apply CV to improving the quality of transportation infrastructure like bus stops and taxi stands.

 
 
Urban Robotics Lab
 
Health monitoring of a large structure using robot system
To monitor a health of a large structure (e.g., bridges), vision and laser based Visually servoed paired structured light system (ViSP) is developed. Kalman filter based displacement prediction algorithm is also developed together with error minimization algorithm for multi-module system.

http://urobot.kaist.ac.kr/projects
 
 
Jellyfish Elimination Robotic Swarm (JEROS)
To cope with rapid increase in jelly fish and related damages, Jellyfish Elimination Robotic Swarm (JEROS) is developed. JEROS is consists of an unmanned surface vehicle (USV) and Jellyfish-shredding module. The robot localizes with GPS and detects jelly fish using visual sensors. The system is expanded to a multi-vehicle system with leader-follower based formation Control.
 
 
Drill direction control system
The research applies robotic sensor technology to localize underground. Directional drilling system is highlighted for economic feasibility and efficiency as it allows drilling to a desired direction underground. We are working on design, control and manufacturing of the drilling robot system by applying state-of-art robot localization technology in extreme environment.
 
 
Autonomous navigation
Localization is a key technology required for autonomous navigation. The research focuses on robotic localization and autonomous navigation. We work on algorithms using 3D Simultaneous Localization and Mapping (SLAM) algorithm, localization, 3D mapping and optimal path planning, and apply them to robots applicable to fields.
 
 
Smart & Sustainable Environmental Laboratory(SSEL)
 
i-Sphere: : A Virtual Reality Based 3D Interactive Web Navigation Interface
The research developed in SSEL explores a novel hybrid interface which dynamically links a virtual reality based 3D hypermedia object with hypertext webpages. The tree structured 3D hypermedia graphic object named I-Sphere consists of multiple nodes and links intuitively represents the 3D site-map of a site so that the user is aware of his/her navigational position in the web-space at all times.

http://ssel.kaist.ac.kr/xe/v20_projects/2314
 
 
ubiSpace: Prototypical smart space built upon swarm intelligence middleware platform

ubiSpace is intelligent digital space system consists of intelligent modules including floor, wall, ceiling. It supports both bidirectional and unidirectional multimedia service that allows virtual travel, conference, and 3D environment awareness.  



http://ssel.kaist.ac.kr/xe/v20_projects/1610
 
 
SEED: Layered agriculture facility using sunlight

SEED enables cultivation of various spices without limitation of climate condition and location. By minimizing energy consumption and CO2 emission, this artificial agriculture prototype outperforms other artificial light based methods (e.g., LED).



http://ssel.kaist.ac.kr/xe/v20_projects/1628
 
 
POSCO square: State-of-art interactive media technology exhibition facility

System design and installation was done for the POSCO SQUARE at KAIST Sports Complex as a new class of interactive multimedia exhibition facility equipped with electrochromic glass, ceiling based full color LED acrylic rods and ultrasonic speakers.



http://ssel.kaist.ac.kr/xe/index.php?mid=v20_projects&page=2&document_srl=1597
 
 
KOSAVE: heuristics-based economical evaluation and decision making system for building Energy

KOSAVE program economically evaluate energy consumption to design and maintain energy efficient building based on heuristics. This system helps users to draw a realistic improvement and solution based on the analysis.



http://ssel.kaist.ac.kr/xe/index.php?mid=v20_projects&category=2092&document_srl=1631
 
 
Smart Transportation System Lab (STSL)
 
Vehicle safety system using V2V communication
This research studies on automatic alert populating system about hazard on road (e.g., vehicle of accident) by using vehicle to vehicle (V2V) communication. This alert system will be more crucial as the automobile IT technology develops with increases in future needs.
 
 
Future traffic prediction algorithm
Various type of sensor information available on the road (e.g., VDS and smart phone) can be used to predict the future traffic. Two approaches, big data-driven and model-driven, are considered according to the objectives
 
 
Real-time railroad safety monitoring platform and big data based maintenance optimization
This research develops overall safety monitoring system for the railroad. This real-time monitoring includes detection of hazard and decision making with prediction. Big data based train maintenance predict the component status and enables effective railroad maintenance.
 
 
Transportation Research and Urban Engineering Lab
 
Safety Management System (SMS) & Risk Assessment
Safety Management System (SMS) aims to provide a proactive approach to system safety of high risk organizations such as airlines, hospitals, and nuclear plants. At TRUE, we created the HRAM (Hazard based Risk Assessment and Modeling), which is a matrix based quantitative risk assessment tool.
 
 
Transportation Analytics
Transportation data has grown exponentially in the past decades, and opened a new channel to analyze and understand numerous common interests in the domain. At TRUE, we utilize multiple data sources for collision analysis and prediction, including feature selection of injury severity prediction, high collision area classification, and post-collision response.
 
 
Air Traffic Flow Management (ATFM)
Air Traffic Flow Management concerns scheduling, planning and control of aircraft at the strategic level. The goal is to maximize the system throughput and on-time performance in the complex air transportation system environment. At TRUE, we focus on probabilistic air traffic management modelling and optimization as well as scenario analysis.
 
 
Transportation and Built Environment
Transportation plays an indispensable role in the build environment, providing means to move people, goods and more. At TRUE, we apply GIS to evaluate the transportation infrastructure reliability in the context of disaster management and prevention.
 
 
Urban Design Lab
 
Urban Design of Experimental, Innovative Approaches
We pursue experimental, innovative approaches in urban design. We seek to design projects by conventional urban design methods, and experimental, innovative design methods that can be tried in academic environments. We also conduct an experimental urban design project considering the future society. In the study on designing the future city where the autonomous vehicle is introduced, the influence of the urban space by the transportation system of the autonomous vehicle was analyzed and the urban space system was proposed accordingly. In Zero Energy Smart Village design research, we proposed a method of village design considering the method of managing energy demand and prediction using artificial neural network.
 
 
Urban Design for Sustainable Environment
We pursue urban design projects to achieve environmental and social sustainability. Based on urban spatial analysis and urban identity analysis, we develop urban design approaches that can identify current characteristics in cities and sustain the characteristics. We carry out urban design projects for sustainable environment such as mixed-use facility design, neighborhood redevelopment, urban regeneration, etc.
 
 
Data-driven Urban Spatial Analysis
We conduct research projects to analyze cities based on data generated in cities, which include urban spatial characteristics, urban perception, urban identity, urban redevelopment impact, and so on. We investigate urban space recognition and urban block classification based on spatial big data, and identify urban characteristics based on social big data.
 
 
Network-based Urban Analysis
We investigate the characteristics of urban space and change of urban structure based on urban network. We also study the effects of urban environment and urban activities due to climate change by building a network that integrates various networks in the city. We analyze the correlation between urban network and various factors that constitute the city, and investigate the characteristics and seek to forecast the change