A method for detecting errors in component-based robotic systems.

This technology is about a method that can detect its own errors using message call logs in a robot system, while reducing false positive detections. It is a method that can be applied to everything from simple cleaning robots to humanoids, and can be applied to service robots implemented as component-based distributed systems to reduce false negative detections due to interfaces that provide limited information in the form of black boxes. It is also a technology that can reduce false negatives due to different scenario execution orders by dividing and recording error detection units by single scenario.

‍

‍

Key Features:

• Reduces the error detection sensitivity of the behavior model of a self-adaptive robotic system.
• Excludes unit functions with an indeterminate call order from behavior model comparison.
• Excludes unit functions with unclear precedence to a specific behavior from comparison.
• Generates message call logs into a single behavior model for each scenario.

‍

The patent is a technology developed through a research project on design and verification technology for collective intelligence-based components for dynamic self-adaptation, which was conducted as a research project for the development of next-generation information and computing technology by the National Research Foundation of Korea.

‍

A technology for constructing a stock portfolio using a program.

This technology collects stock price data for each stock, analyzes correlations between the collected data, and predicts the stock prices of the correlated stocks using a stock price prediction model learned from the stock price data. The predicted prices are then compared to select stocks with a predicted increase in price from the analyzed correlations, thereby increasing returns.

‍

Key Features:

• Stock portfolio construction using deep learning techniques
• Investment techniques for diversifying multiple stocks
• Analyzing correlations between stocks from stock price data
• Predicting stock prices using a prediction model

‍

The patent is a technology developed with research funding from the Ministry of Education on "Integrated Processing and Analysis of Streaming and Stored Data in a Distributed Environment."

Active Noise Cancelling Technology.

This technology utilizes a three-level digital signal to reduce loop delay, either using a three-level PWM (pulse width modulation) signal or a three-level PDM (pulse density modulation) signal. This reduces loop delay, thereby increasing the frequency range for active noise reduction in earphones receiving digital audio signals.

‍

Key Features :

• Uses a three-level digital signal.
• Attenuation is achieved by reducing loop delay.
• This effect increases the peak frequency of noise signals from 500 Hz to over 2,500 Hz.
• This allows you to eliminate ambient noise in subways and airports, allowing you to hear only the desired audio signal.

‍

A technology for manufacturing a plant-based milk substitute using oats and almonds as the main ingredients.

This technology enables the commercial production of plant-based milk using oats and almonds, with the same taste, texture, and nutritional content as cow's milk. This manufacturing method can be used as a milk substitute for those with lactose intolerance, contributing to public health.

Key Features:

• A high-quality beverage using lecithin as an emulsifier.
• Controls viscosity changes and ensures colloidal stability.
• Enzyme treatment of oats.

‍

Neuroscience image processing technology.

Technology for determining the location of a DBS (Deep Brain Stimulation) electrode. This technology involves aligning pre-implantation MRI volume data sets with post-implantation CT volume data sets to determine the location of a subvolume-rendered DBS electrode relative to the STN (subthalamic nucleus).

‍

Key Features:

• Deep brain stimulation.
• DBS treatment.
• Stereotactic neurosurgery.
• Image synthesis technology using volume rendering.

‍

This patent is a German family patent for the "Method for Determining the Location of a DBS Electrode" of Korean patent.

Neuroscience image processing technology.

Technology for determining the location of a DBS (Deep Brain Stimulation) electrode. This technology involves aligning pre-implantation MRI volume data sets with post-implantation CT volume data sets to determine the location of a subvolume-rendered DBS electrode relative to the STN (subthalamic nucleus).

‍

Key Features:

• Deep brain stimulation.
• DBS treatment.
• Stereotactic neurosurgery.
• Image synthesis technology using volume rendering.

‍

This patent is a U.S.A family patent for the "Method for Determining the Location of a DBS Electrode" of Korean patent.

Neuroscience image processing technology.

Technology for determining the location of a DBS (Deep Brain Stimulation) electrode. This technology involves aligning pre-implantation MRI volume data sets with post-implantation CT volume data sets to determine the location of a subvolume-rendered DBS electrode relative to the STN (subthalamic nucleus).

‍

Key Features:

• Deep brain stimulation.
• DBS treatment.
• Stereotactic neurosurgery.
• Image synthesis technology using volume rendering.

‍