Sonar Image Training Data Generation Technology Using Ray Tracing Simulation and Measured Noise Synthesis

This technology is a sonar image simulator that generates virtual sonar images using a 3D model-based ray tracing method, combines them with background noise images captured in actual underwater environments to create training images, and trains models to detect underwater objects.

Previously, it was difficult to secure large-scale sonar image datasets for training neural networks for underwater object recognition. Furthermore, collecting real-world data was time-consuming and costly, and discrepancies between simulated images and real-world environmental noise often led to degraded recognition performance.

This technology proposes a method to generate precise training data by simulating sonar images using 3D object models and sample rays, then synthesizing them with measured backgrounds to which Gaussian blur and noise level parameters have been applied. Applicable to underwater exploration, marine disaster prevention, and port surveillance, it significantly reduces data collection costs while enhancing detection performance.

‍

‍

Key Features:

• A simulator that generates sonar images of virtual objects based on ray tracing
• An imaging unit that captures actual underwater backgrounds to generate background noise images for training
• A synthesis unit that combines simulated sonar images with noise images to create training data
• A training unit that builds models for detecting underwater objects by training on the generated images

‍

This invention was developed with support from the Smart Underwater Tunnel System Research Center of the Ministry of Science and ICT.

Task Allocation Technology for Two Heterogeneous Robots Based on Primal-Dual Heuristics

This technology is a task allocation method and system that applies primal-dual heuristics to distribute tasks between two structurally heterogeneous robots, iteratively adjusting cost weights for each robot to minimize the maximum total travel cost.

Existing challenges included the difficulty of achieving efficient task allocation and path planning to balance task completion times and minimize the maximum travel cost of the entire system in environments where the two robots have different performance capabilities.

This technology proposes a method that assigns weights to each robot's travel cost, distributes destinations using primal-dual heuristics, and then adjusts the weights to reduce the higher value by comparing the calculated total travel costs to derive an optimal path. It provides a practical solution for significantly reducing total task completion time in environments where robots with different performance levels coexist, such as logistics warehouses, factory automation, and disaster prevention patrols.

‍

‍

Key Features:

• Weight registration step where cost weights for each heterogeneous robot are registered
• Destination allocation step where multiple destinations are distributed to each robot using primal-dual heuristics
• Path planning step where the allocated destinations are applied to a path planning algorithm to calculate travel routes and total travel costs
• Configuration that iteratively adjusts cost weights to reduce the higher value among the total travel costs

‍

This invention was developed with support from the National Research Foundation of Korea for the Intelligent Growth Autonomous Driving System for Unmanned Vehicles Operating Safely in Congested Residential Road Environments.

Upper Limb Assistive System for Throwing Motions Combining Torsion Springs and Functional Electrical Stimulation

This technology is an upper limb assistive system that supports arm flexion and extension through a brace worn on the user's arm and a torsion spring-based hinge structure, while inducing active contraction of wrist and arm muscles via functional electrical stimulation to perform throwing motions.

Rehabilitation programs for patients with upper limb impairment due to conditions like stroke have traditionally relied on passive, repetitive electrical stimulation, often resulting in low patient engagement and limited therapeutic outcomes.

This technology integrates a latch-based trigger unit that locks and releases the arm's folded state with a torsion spring that provides elastic restorative force. A control unit synchronizes the timing of electrical stimulation with the trigger release, assisting the patient in performing an actual throwing motion. By being applicable to both stroke rehabilitation and sports-oriented therapy, it encourages active patient participation and interest, significantly enhancing rehabilitation effectiveness.

‍

‍

Key Features:

• Multiple wrist stimulation electrodes that provide electrical stimulation to muscles involved in wrist extension and flexion
• First and second braces for the upper and lower arm, respectively, connected by a hinge axis that allows for joint movement
• A torsion spring installed on the hinge axis to move the braces from a folded to an extended state using elastic force
• A control unit that synchronizes the timing of electrical stimulation and trigger release to assist in throwing motions

‍

This invention was developed with support from the Ministry of Culture, Sports and Tourism’s project for hybrid smart clothing and monitoring systems for athletic performance enhancement, and the Ministry of Science, ICT and Future Planning’s Human-Centered Soft Robot Technology Research Center.

Robotic Bronchoscopy Technology Using a Remote-Controlled AI Suction Device

This technology is a robotic bronchoscopy system and control method using an AI-powered suction device that remotely controls a bronchoscope camera and suction catheter to access peripheral bronchi for specimen collection and mucus aspiration.

Conventional methods are highly dependent on the operator's skill level, making immediate intervention difficult in emergencies and lacking remote operation capabilities.

This technology proposes a method to remotely control a robotic device—comprising imaging and suction units—via an interface, while variably adjusting screen displays and drive inputs based on camera positioning. This enables stable procedures regardless of operator skill. It can be applied in environments where medical staff contact must be minimized, such as intensive care units and infectious disease wards, thereby enhancing both procedural safety and medical accessibility.

‍

‍

Key Features:

• An imaging unit including a camera that captures real-time internal images of the patient's bronchial area.
• A suction unit including a suction catheter inserted into the bronchi to collect specimens.
• An interface unit that displays first-state information linked to camera operation and second-state information linked to suction catheter operation.
• A configuration that switches the interface based on pressure data and insertion depth information applied to the suction catheter.

Wearable Robotic Glove for Grip Strength Enhancement Based on Single EMG Sensor MAV Analysis

This technology is a control system for a wearable robotic glove that assists with grip strength by placing a single electromyography (EMG) sensor at the musculotendinous junction of the flexor digitorum superficialis to measure the Mean Absolute Value (MAV) of the EMG signal and identify the user's intent in real time.

Conventional multi-sensor EMG analysis methods require individual user training, precise sensor placement, and are prone to frequent malfunctions caused by body movement.

This technology proposes a method of controlling the glove's grip module by placing a single EMG sensor at the musculotendinous junction of the flexor digitorum superficialis and triggering the device when the measured MAV falls outside of user-defined activation and deactivation thresholds. This allows for stable grip enhancement with a simple configuration that eliminates the need for individual training. It can be used to assist workers in industrial settings and support daily activities for the elderly and patients, while the single-sensor design reduces production costs, accelerating the mass adoption of wearable robots.

‍

‍

Key Features:

• A sensor module worn on the user's forearm, a control module worn on the body, and a grip module worn on the hand
• A sensor band configuration that precisely positions the sensor unit at the musculotendinous junction of the flexor digitorum superficialis, which is responsible for finger flexion
• A sensor unit consisting of a single EMG sensor that captures EMG signals from the musculotendinous junction of the flexor digitorum superficialis
• An identification algorithm that detects grip intent and controls the grip module based on the MAV, calculated by summing the absolute amplitudes of the captured EMG signals

‍

This invention was developed with support from the Human-Centered Soft Robotics Research Center of the Ministry of Science and ICT.

Modular aircraft with thrust vector control via a 2-axis joint and coupling capability

This technology features a flight module and an aircraft equipped with it, where the thrust unit is connected to the interior of the flight module's body via a 2-axis joint to enable roll and pitch movements, allowing multiple modules to be combined to form variable aircraft configurations.

Existing multi-rotor aircraft face challenges in adapting to diverse missions due to takeoff weight limitations caused by thrust constraints, short flight times, and restrictions on the type and performance of onboard equipment.

This technology proposes a method to increase thrust and extend flight time by combining, separating, and rearranging multiple flight modules during flight. By using joints and actuators to control the thrust direction of each module, flight efficiency can be optimized. In industrial drone sectors requiring cargo transport, large structure inspection, and mission-specific configurations, this system dramatically increases operational flexibility by allowing performance to be scaled simply through module coupling.

‍

‍

Key Features:

• A body with an interior space open at the bottom, and a thrust unit that moves within the interior space to generate thrust
• A joint that connects the thrust unit to the body to enable roll and pitch movements
• An actuator that drives the joint to rotate the thrust unit relative to the body, thereby controlling the direction of thrust
• The joint includes a bracket formed by coupling a first plate and a second plate that are orthogonal to each other, along with first and second rotating disks

Hip Joint Linkage for Wearable Robots Generating Multi-Directional Gait Moments via a Four-Bar Mechanism

This technology is a hip joint linkage for wearable robots that utilizes a four-bar mechanism consisting of multiple rotary joints and links to convert input torque from a single actuator into multi-directional output moments—including flexion, extension, abduction, and adduction—synchronized with the gait cycle.

Existing hip joint modules for wearable robots typically require multiple actuators to assist with multi-axial moments, leading to increased system weight and complexity.

This technology proposes a configuration where the drive unit is positioned at the first rotary joint, utilizing a four-bar linkage mechanism composed of six rotary joints and four links. By shifting the instantaneous axis of rotation, it can generate the appropriate output moment direction for each phase of the gait cycle—such as extension and abduction during the early phase, abduction during the mid-phase, and flexion and abduction during the late phase—using only a single actuator. Applicable to gait assistance, industrial strength support, and rehabilitation training, this solution significantly contributes to the lightweight design and cost reduction of wearable robots by enabling multi-directional assistance with a single motor.

‍

‍

Key Features:

• A hip joint unit connecting the upper body link, which is fixed to the upper body of the wearable robot, to the femoral link connected to the thigh.
• The hip joint unit is configured with multiple rotary joints and links connected as a four-bar mechanism.
• The input moment is converted into output moments in specific directions depending on how it is transmitted through the rotary joints and links.
• Outputs moments in the direction of extension and abduction during the early gait phase, abduction during the mid-phase, and flexion and abduction during the late phase.

Length-sensing sensor that measures artificial muscle deformation via resistance changes in conductive knitted fabric

This technology features an elastic knitted layer that wraps around the outer surface of an artificial muscle. By knitting conductive yarn into this layer, the sensor measures changes in the yarn's resistance as the artificial muscle expands or contracts, thereby detecting its deformation.

Conventional microchannel-based EGaIn sensors suffer from channel deformation under high-pressure conditions, which hinders accurate contraction prediction and reduces durability.

This technology utilizes a highly elastic knitted layer with conductive yarn integrated via plain or rib knitting. It prevents separation between the artificial muscle and the sensor layer using fixing loops and clamps, while a braided mesh ensures the conductive band structure remains stable. Suitable for wearable robots, soft actuators, and rehabilitation aids, it ensures accurate contraction prediction and enhanced durability even in high-pressure environments.

‍

‍

Key Features:

• A knitted layer with interlocking yarn loops that surrounds the outer perimeter of the artificial muscle expansion member
• Fixing loops knitted into the fabric layer to provide conductivity and secure both ends of the layer to the artificial muscle
• Clamps pressure-fitted onto the outer surface of the fixing loops to prevent separation between the artificial muscle and the knitted layer
• A braided mesh that surrounds the outer perimeter of the conductive fabric or band and is positioned between the fixing loops and the clamps

‍

This invention was developed with support from the Ministry of Science and ICT for the development of biomimetic bionic arm mechanisms.

Multi-DOF Joint Mechanism with String Sets for Shock Absorption and Posture Restoration

This technology implements a multi-degree-of-freedom (DOF) joint mechanism that uses string sets to minimize mechanical interference between joint components, absorb external shocks, and restore the initial posture.

Conventional robot arm joints suffer from user discomfort due to the transmission of external shocks caused by their rigid structures and the weight burden when worn as prosthetic limbs.

This technology connects joint components with string sets to reduce weight and allows each component to rotate while spaced at a predetermined distance, providing shock absorption and restorative force. It can be applied to robotic prosthetics, wearable robots, and collaborative robots to minimize external shock transmission while enhancing durability and comfort.

‍

‍

Key Features:

• A first joint component that rotatably supports a 2-1 joint component configured to rotate in a first direction
• A 2-2 joint component configured to intersect with the 2-1 joint component as it rotates in a second direction relative to the 2-1 joint component
• A first string set connecting the first joint component and the 2-1 joint component to provide a first degree of rotational freedom
• A second string set connecting the first and 2-1 joint components to the 2-2 joint component to provide a second degree of rotational freedom

‍

This invention was developed with support from the Ministry of Science and ICT for the development of biomimetic bionic arm mechanisms.

Tendon device for wearable robotic clothing to assist muscle strength

This technology is a mechanical mechanism that assists in the flexion and extension of human joints. It features a pair of wires and tendon modules with built-in individual torsion springs, enabling precise muscle strength assistance for each joint by calculating wire tension through torsion spring displacement measured by pulley encoders.

Conventional technology uses a single pulley for symmetrical joint movement, which leads to wire length imbalances. This makes it difficult to provide muscle support during irregular movements such as walking on stairs or inclines and limits control due to the inability to measure wire tension.

This technology applies independent tendon modules for each wire and installs torsion springs inside the pulleys to maintain wire tension and ensure back-drivability. It performs precise control by measuring human-robot interaction forces in real-time using the rotation angles calculated by pulley encoders and the spring constants. It can be applied to rehabilitation training, gait assistance, and medical/welfare services, improving precise control over joint movement and enhancing muscle strength support for daily activities.

‍

‍

Key Features:

• A control unit that outputs control signals by calculating signals generated by the movement of the first and second wires in the first and second tendon modules, based on the tension of the first and second wires measured by the first and second pulley encoders.
• First and second wires that are fixed to the anterior and posterior sides of the user's joint, respectively, and move longitudinally according to the flexion and extension of the joint.
• A drive unit around which either the first or second wire is wound to assist the muscle strength of the agonist muscle during joint movement.
• A tendon device for wearable robotic clothing that provides muscle strength assistance, where the torsion of the second torsion spring maintains the tension of the second wire while providing back-drivability.

‍

This invention was developed with the support of the Ministry of Trade, Industry and Energy for the development of a wearable robotic system consisting of a 50W-class drive module for human muscle strength assistance and a human-robot muscle model-based control technique.

Control System for Human-on-a-Chip

This technology utilizes multiple magnetic field coils to apply independent external magnetic fields to the entire or local areas of a human-on-a-chip, enabling the actuation (fluid flow control) and precise positioning (organ docking) of microrobots injected inside.

Existing human-on-a-chip systems rely on constant flow control using external pumps, which makes local precision control difficult and results in complex structural designs and low efficiency when docking artificial or micro-organs at specific locations.

This technology uses independently arranged magnetic field coil units around the chip network to perform local/global magnetic field steering and intensity modulation. By controlling the rotational speed and direction of helical or propeller-type microrobots, it simulates fluid flow or precisely docks microrobots at target locations. Applicable to industrial robots and automation systems, it improves the efficiency of micro-organ positioning and fluid flow within the network, enabling the creation of environments that closely mimic the human body.

‍

‍

Key Features:

• Local area magnetic field coil unit that applies an external magnetic field locally to areas of the biological organ model and network.
• The magnetic field coil device includes a global area magnetic field coil unit that applies an external magnetic field across the entire chip area.
• The global area magnetic field coil unit controls the positions of multiple microrobots simultaneously to collectively place them at pre-determined micro-organ locations within the human-on-a-chip control system.
• A network connecting the micro-organs that constitute the biological organ model.

‍

This invention was developed with support from the Ministry of Science, ICT and Future Planning for the development of 3D precision magnetic microstructures and cell/drug delivery-based technology.

Pneumatic Artificial Muscle with Enhanced Response Speed Using Elastic Energy

This technology relates to a pneumatic artificial muscle unit and its operating method that utilizes elastic energy. By incorporating passive elastic elements, it achieves rapid contraction and relaxation movements.

Conventional pneumatic artificial muscles suffer from slow response speeds and limited contraction rates due to the time required for pressurization and exhaust, which significantly hinders real-time motion assistance in wearable and collaborative robots.

This technology combines an inelastic sleeve and an elastic tube with a tension spring to store and release elastic energy, thereby improving response speed and enabling rapid contraction and expansion.

‍

‍

Key Features:

• An inelastic sleeve that forms the pneumatic artificial muscle unit and performs repeated contraction and expansion movements along its longitudinal axis.
• A gas inlet connected to one end of the inelastic sleeve for gas injection, and an exhaust valve connected to the other end for gas discharge.
• A tension spring and an elastic tube positioned between the gas inlet and the exhaust valve within the inelastic sleeve to accumulate elastic energy.
• A trigger mechanism that opens the opening/closing member within the exhaust valve when the inelastic sleeve reaches a predetermined expansion length.

‍

This invention was developed with support from the Ministry of Trade, Industry and Energy for the development of a human-augmentation hybrid robot suit capable of running 100m in 7 seconds and being worn comfortably for 12 hours, and from the Ministry of Science and ICT for the development of core technologies for human-robot interaction-based hybrid control and interface design for safe and efficient collaboration, mobility, and rehabilitation.

Multi-directional Sensing Wearable Device for Human Motion Assistance

This technology relates to a wearable device for assisting human motion, which detects multi-directional movement and provides assistive force through sensors composed of internal/external electrodes and deformation elements.

Existing wearable sensors could only detect deformation in a single direction, making it difficult to accurately capture complex joint movements, and achieving multi-directional sensing often led to complex configurations and increased costs.

This technology implements multi-directional sensing with a simplified electrode structure, and an actuator provides assistive force based on the sensing results to support human motion. It can be widely applied to robotic hands, gait assistance devices, and more.

‍

‍

Key Features:

• An inner electrode positioned at the center of a circle and a plurality of outer electrodes arranged at equal intervals along the circumference
• A plurality of deformation elements disposed between the inner electrode and each outer electrode, made of a material whose resistance changes upon deformation
• A control device that applies different control signals to an actuator based on the combination of different resistance values from each deformation element
• An actuator that assists the wearer's physical movement by operating based on the different control signals applied by the control device

‍

This invention was developed with support from the Ministry of Science and ICT for self-powered multifunctional fabric manufacturing technology.

Multi-legged walking robot with pneumatic support for stance phase

This technology features a multi-legged walking robot equipped with multiple independently driven leg units. Pneumatic devices connected to the hip and knee joints of each leg unit assist the actuator torque required for ground support during the stance phase and disengage during the swing phase.

Multi-legged robots typically require high torque during the stance phase, which places excessive load on the actuators. This hinders rapid leg movement during high-speed locomotion and reduces overall energy efficiency.

This technology proposes a method to toggle the pneumatic assistance based on the stance and swing phases, allowing the actuators to drive the joints more efficiently. It can be applied to high-speed robots, rough-terrain exploration, and military robot platforms, significantly reducing the burden on actuators while improving both speed and energy efficiency.

‍

‍

Key Features:

• Includes multiple independently driven leg units, with each leg unit featuring a hip joint driven by an actuator.
• Includes a femoral member connected at one end to the hip joint, which rotates in response to the hip joint's movement.
• A knee joint connected to the other end of the femoral member, and a lower leg member connected at one end to the knee joint.
• Pneumatic devices mounted on the hip and knee joints to provide support during the leg unit's stance phase.

‍

This invention was developed with support from the Ministry of Science, ICT and Future Planning for research into new hybrid actuators for high-speed robots.

Underwater Tunnel Module Joining Monitoring System Combining Sonar and Line Laser Scanning

This technology is an underwater monitoring system that identifies the relative position between an existing tunnel module and a connecting module using sonar, and calculates the gap, height difference, and 3D imagery of the joint using a mobile underwater scanning device equipped with a line laser projector and an underwater camera.

In underwater environments, GPS is unavailable and visibility is low, which limits the use of optical sensors. Furthermore, the difficulty of deploying divers has historically made precise positioning and guidance for tunnel module joining a significant challenge.

This technology proposes a method that first verifies the distance using sonar, then performs a precise scan of the joint with an underwater scanning device to compare the 3D coordinate endpoints of the surface projected by the line laser, thereby deriving the gap and height difference. Applicable to the construction of underwater tunnels and subsea structures, it enables precise joining without the need for divers, significantly enhancing construction safety and accuracy.

‍

‍

Key Features:

• Sonar device that identifies the relative position between the existing tunnel module and the connecting tunnel module within the sonar imaging range
• Underwater camera that captures images of the area where the two tunnel modules are to be joined
• Laser projector that casts a line laser onto the existing and connecting tunnel modules
• System configuration that calculates the joint gap and height difference by comparing the 3D coordinates of the surface projected by the line laser

‍

This invention was developed with support from the Smart Underwater Tunnel System Research Center of the Ministry of Science and ICT.