Bidirectional Protocol at Wireless Sensor Line Protocol

• BiWSLP(Bidirectional Wireless Sensor Line Protocol)
• Adaptive BiWSLP

  - Supports data transfer in the upward direction (sensor node -> management server) and downward direction (management server -> sensor node) in multi-hop environment

  - Secures real-time data using time-divisional (TDMA)

  - Controls sub-real-time sensor node through down direction command

• Field of Application

  - Wireless monitoring systems

  - Wireless alarm and detecting systems

  - Home Automation

Image Transmission Network Technology

• WITP (Wireless Image Transmission Protocol) Technology

  - Controls data errors and flow

• WIRP (Wireless Image Routing Protocol) Technology

  - Eliminates bottleneck to gurantee shortest path algorithms protocals

MOTA Programming (Multi-hop Over The Air Programming)

• Remote upgrade of node & gateway software
• Firmware update of sensor node through wireless transmission of firmware from central server to remote site
• Reduces management cost required for firmware upgrade

  - IR Sensor Modules

  - Temperature/Humidity Sensor Modules

  - Smoke Detection Sensor Modules

  - Flame Detection Sensor Modules

  - LVDT Type Movement Sensor Modules

Low Power Technology

• Low-Power, High-Reliability Network Protocol Technology
• Low-Power Sensor Nodes & Gateway Development Technology

Gateway Technology

• Wireless Internet (CDMA/HSDPA) Communication with a server
• Remote Sensor Node Control & Management Technology
• Solar Cell Charging Circuits & Low Power Operating Technology

Gateway Techonology

• Open source hardware operating technology
• Interworking technology of heterogeneous IoT devices
• Various sensor interface technology
• Various wireless interworking technologies (BLE, LAN, Wi-Fi, RF(2.4G, 900M, 477M)
• Provides know-how of installation and operation of various servers (Media, DB, WEB, FTP, NAS)
• Supports International IoT standard (OneM2M (MQTT)

Real Time Remote Monitoring Applications

• Remote fire monitoring and detecting technoloy
• Remote control technology
• Acquisition and analysis technology of Sensing data

Integrated Operation and Management Service Platform for the AI Lifecycle (D-Lab Flow)

• A comprehensive platform that integrates all stages of AI learning, including data collection, labeling, model training and optimization, deployment, and management
• Function

  - Providing data collection and management through IoT device integration, as well as preprocessing functions such as image cropping

  - Providing dataset creation and management, labeling collaboration, object detection class definition, various labeling tasks, and auto-annotation features

  - Providing IoT device deployment for AI models, algorithm downloads, REST API integration, and system monitoring capabilities

  - Providing AI model optimization, version control, image preprocessing and augmentation, and training monitoring features through MLOps integration

  

MEDBIZ Health Care Analysis Platform

• Customized, high-quality medical service platform through medical information collection (IoT-based) and health information analysis (big data-

  

• Function

  - Provides acquisition and management of IoT device and existing IoT system data

  - Supplies managemnet of virtual file system based IoT data

  - Provides managemnet and monitoring of platform using a website

  - 4P(Predictive, Preventive, Paticipatory, Presonalized)Personalized disease prevention services based on

  - Big data analysis function linked with analysis technology of Hadoop, R etc

  - Offers encryption technology and data encryption function

  - Provides Open API to develop various application services

Smart Tourism Big Data Platform

• A platform analyzes data that is collected through smartphones and IoT devices and provides tailored services to tourists in conjunction with public data

  

  

Distributed computing platform based on IoT virtualization

• A platform capable of storing and managing heterogeneous IoT data provides cloud storage and utilizes a distributed computing system for large data analysis

  

• Function

  - Collecting IoT device and existing IoT system data and providing integrated management functions

  - Provides virtual file system-based IoT data management

  - Provides platform monitoring and management using the Web

  - Works with Hadoop, R and others to provide big data analytics

  - Provides encryption technology and data encryption

  - Open API to develop a variety of application services

NILM(Non Intrusive Load Monitoring)

• System to monitor power without disturbing existing environment
• Comparison of electricity usage with other houses and set-top box monitoring function

  

MEDBIZ Healthcare Analysis and Service of Web Development

• A Web service enables users to utilize cloud functions such as virtual file system, device information, big data analysis provided by the platform through the UI of a web browser

  

Data analysis environment Middleware-based greenhouse growth environment analysis / control

• It gathers the various greenhouse environment information and finds the most suitable enviornment for the growth of crop

  

Data Analysis Services

• Provides analysis and prediction of diseases based on statistical and mathematical techniques for patterned data analysis as well as information on high frequency therapy collected using pattern recognition

  

Location-based services

• Services that provide location of people on vehicles or bicycles through information via mobile wireless network

Building MLOps for D-lab Technology Development

[K3s Cluster & Client Setup Based on Kubernetes]

• Kubernetes serves as a service for managing users in D-lab Flow, providing functions like resource allocation, control, scaling, and monitoring to enhance the operational efficiency of the system

  

  

[Kubeflow Setup Based on Kubernetes]

• A service for managing and automating AI workflows (data preprocessing, model training, model evaluation, etc.), which runs smoothly in a Kubernetes environment and integrates with multiple AI frameworks (TensorFlow, PyTorch, etc.) to support experiment tracking and pipeline management
• From an operator’s perspective, this service helps manage AI models used by users, supporting experiment tracking, pipeline management, and automation

  

  

[MinIO Setup Based on Kubernetes]

• Images stored in D-lab Flow are saved in MinIO (a distributed object storage service), supporting efficient storage and fast access to AI model datasets and training results

  

[BentoML Setup Based on Kubernetes]

• Supports efficient packaging and deployment of AI models for preprocessing, training, and inference
• Useful for deploying and managing services between Kubernetes and frontend

  

Integrated Labeling Technology for Consistent Annotations

• Applied integrated labeling technology (with consistent standards) to prevent model performance degradation and training failures

  

White Blood Cell Detection Technology in Blood

• Using Object Detection technology to assist in diagnosing blood diseases and developing automated medical analysis systems

  

Low-Resolution Image Correction Technology

• AI algorithm technology removes noise from CT images to improve image quality, thereby helping to improve diagnostic accuracy

  

Keyword-Based Literature Information Analysis System Development

• An automated system that retrieves papers/patents, trend data, etc., based on input keywords and outputs results such as AI model analysis and graphs in PDF format
• An automatic report generation system for systematic document analysis and screening automation using large-scale language models

  

Preprocessing Methods for Imbalanced Data and Optimal Learning Model Recommendation System

• A system for detecting the optimal learning model according to methods for addressing data imbalance
• Developed a system for generating synthetic data based on structured data, considering statistical characteristics and class imbalance issues in financial data to improve algorithm performance

  

  

Wildfire Synthetic Data Technology Development

• Using Stable-diffusion technology for generating synthetic data based on images by inputting background images and text to create wildfires, with testing for control
• Improving the quality of synthetic wildfire data to aid in training

  

Automated Nuclear Boundary Detection Algorithm for H&E Stained Thyroid Tissue Images

• Utilizes morphological and statistical methods to identify the center of the nuclei, count the number of nuclei, and detect their boundaries

  

Machine Learning-Based Quantitative Analysis Algorithm for Immunohistochemistry Stained Tissue Images

• Developed an algorithm that quickly performs quantitative analysis of antibody reaction areas based on machine learning, providing both the overall image and the nucleus count analysis, expressed as grades and ratios

  

Certificates

1. 

   - MEDBIZ Platform

   - Software Quality Certification (GS) Grade 1

Award

1. 

   - The World Intellectual Property Organization (WIPO) Selected as a Successful Case of Patent Technology Commercialization

2. 

   - Sponsored by the Patent Office Gangwon 10th invention prize

3. 

   - The Ministry of Knowledge Economy Local SW competition The National Institute for Information and Communication Development Award

4. 

   - The Ministry of Education, Science and Technology Commendation of Ministry of Industry-University Cooperation

5. 

   - The Ministry of Trade, Industry and Energy Commendation for the Minister of Employment

6. 

   - The Ministry of Science, ICT and Future Planning ICT Innovation Award The Dong-A Daily News Award