Multi-legged modular robot with a multifunctional end-effector and coupling interface

This technology is a control system for multi-legged modular robots that can be coupled or decoupled. It identifies idle legs when robots are connected and drives the hinge of the leg to selectively expose either a contact tip (for walking) or a gripper tip (for manipulation), allowing for variable functionality of the legs.

When multi-legged robots are combined into a swarm, the number of legs increases significantly; however, the inability to efficiently utilize idle legs limits the overall functional scalability of the robot.

This technology stores information regarding the coupling interface and the overall structure of the combined robots in memory. It automatically detects idle legs based on the coupling state and uses a control algorithm to drive the hinge of the multifunctional end-effector, selectively exposing either the gripper tip or the contact tip to switch between manipulation and walking functions.

‍

‍

Key Features:

• A first leg unit consisting of a first main body and first through fourth legs that support the load by contacting the ground
• A first coupling interface for connecting with a second multi-legged modular robot, a first coupling unit, and a first mechanical drive unit that controls the coupling and leg operations
• Multiple joints coupled between connecting rod sections that adjust the position of the multifunctional end-effector to perform either walking or manipulation functions
• A multifunctional end-effector comprising a tip section with a contact tip at one end and a gripper tip at the other, along with an external housing case

‍

System for controlling mobile robots using code blocks and projection visualization

This technology is a cyber-physical system (CPS)-based control solution that operates physical mobile robots using executable code generated from user-inputted code blocks. It utilizes a projection module to overlay virtual scenario objects onto the physical driving environment, providing real-time visualization and feedback.

Conventional code block-based educational content is limited to virtual environments, making it difficult for users to experience the interaction between the physical world and digital content, which restricts the development of practical robot application skills in real-world settings.

This technology implements integrated virtual-physical interaction through a system architecture that includes real-time positioning of physical robots using 3D depth sensors, visualization of environmental images and scenario objects via a projection module, and the conversion of user-assembled code blocks into robot executable code, along with location-based situational awareness and feedback generation.

‍

‍

Key Features:

• A projection module that visualizes and displays scenario-related objects by projecting them onto the mobile robot's driving environment.
• A user code block generation module that creates code blocks to control the mobile robot's actions based on user-assembled inputs.
• A code block conversion module that translates the generated code blocks into executable code for controlling the mobile robot.
• A mobile robot control module that executes the code to manage movement based on the robot's position and surrounding situational awareness data.

‍

Reinforcement Learning-Based Gait Generation Technology Using Reward Functions Without Reference Motions

This technology combines deep reinforcement learning with finite state machines to generate character gait motions in real-time. It inputs dynamic states and character-specific parameters into a neural network to generate action information, learning natural gait policies through reward functions.

Conventional finite state machine-based control methods have limitations in achieving natural motion, while existing deep learning approaches require separate reference motion data for gait decision-making, reducing their versatility and efficiency.

By incorporating the minimization of gait parameter deviation, vertical axis maintenance, directional alignment, and joint torque minimization into the reward function, this technology determines optimal stance hip torque and joint angles without the need for reference motion data. It can be applied to bipedal robot control and the generation of character motions in games and animation, enabling natural gait implementation without the burden of data collection.

‍

‍

Key Features:

• A step of a computing device acquiring gait motion state information of a character at time t
• A step of inputting the acquired state information into a pre-built neural network model to generate action information
• A step of transmitting the action information at time t to the character or a device that outputs the character's motion
• A configuration where the neural network model receives gait motion states as input to determine action information that maximizes reinforcement learning rewards

‍

This invention was developed with support from the Ministry of Science and ICT for the development of biomechanical model-based intelligent control technology for human movement involving multi-level interactions, and the DeepXR: Deep Hyper-Reality research project.

Four-bar linkage-based stair-climbing device with a flexible fit structure

This technology is an obstacle-climbing device that uses a four-bar linkage mechanism to drive rotating legs to overcome obstacles. It features a flexible fit structure on the bottom of the legs, allowing for variable contact area and reaction force depending on the dimensions of the stairs.

Existing crank-leg or tracked robots are designed for specific stair dimensions, which limits their ability to navigate stairs of varying sizes.

This technology incorporates a flexible fit on the bottom of the legs, featuring multiple supporters spaced between the upper and lower bases. The spacing and tilt angles of these supporters are differentiated by region, allowing them to deform variably upon contact. It can be applied to indoor delivery robots and disaster response robots, ensuring stable climbing performance even in environments with irregular stair dimensions.

‍

‍

Key Features:

• A pair of rotating legs located on both sides of the body, capable of relative rotation with respect to the body
• A support leg provided on the bottom of the body and arranged parallel to the rotating legs, and a leg drive unit that rotates the pair of rotating legs
• A flexible fit provided on the bottom of at least one of the rotating legs or the support leg
• A plurality of supporters connecting the upper base and the lower base, arranged at predetermined intervals

Rehabilitation Robot Control Device

This technology measures a rehabilitation robot user's brain signals (specifically changes in blood flow) using functional near-infrared spectroscopy (fNIRs) and compares them against machine learning-based pain patterns to determine the presence and intensity of pain. It then uses this data as a control logic to automatically adjust the robot's operating intensity or trigger an emergency stop.

Conventional manual emergency stop buttons are difficult for patients to press in an emergency, and existing physical quantity sensing methods have limitations in accurately responding in real-time to pain outside the training range or sudden situational changes.

This control device consists of a sensor unit that monitors the user's cerebral blood flow, a processing unit that recognizes pain patterns, and a control unit that automatically stops the robot or adjusts its intensity based on pain signal trends when the signals exceed a preset threshold. Applicable to rehabilitation training, gait assistance, and medical/welfare services, it enhances the safety and effectiveness of rehabilitation by automatically adjusting robot operating intensity based on the user's brain signals.

‍

‍

Key Features:

• A brain signal processing unit that determines whether the measured brain signals indicate pain by comparing them against preset pain patterns.
• A rehabilitation robot control device including a rehabilitation robot control unit that regulates the robot's operating intensity based on whether the measured brain signals indicate pain.
• A brain signal measurement unit that measures brain signals corresponding to changes in cerebral blood flow from a user utilizing a rehabilitation robot.
• A brain signal processing unit that determines whether the measured brain signals indicate pain.

‍

This invention was developed with support from the Ministry of Education, Science and Technology for the development of upper-limb rehabilitation robot technology using EXG for cognitive/motor rehabilitation of patients with upper-limb paralysis.

Surgical Robot

This technology is a surgical robot mechanism that utilizes a leaf spring-type drive transmission unit within a multi-joint positioning unit. By adjusting the bending and straightening of the leaf spring based on the tensile force of the drive unit, it controls joint positioning and generates high driving torque.

Conventional wire-based drive transmission methods struggle to provide the sufficient torque required for manipulating internal organs and face physical limitations regarding the surgical workspace and power transmission accuracy.

Instead of wires, this technology places a leaf spring-type drive transmission unit on one side of the joint and controls it via a drive unit, increasing mechanical rigidity to deliver high torque. A guide member ensures precise, stable operation without displacement. Applicable to surgical robots, interventional systems, and medical automation, it improves procedural accuracy and ensures precise drive transmission, thereby reducing the burden on the patient.

‍

‍

Key Features:

• A surgical unit coupled to the positioning unit, which moves in tandem with the positioning unit and features a surgical tip at the distal end for performing procedures on the affected area.
• A leaf spring-type drive transmission unit that controls the position of one joint relative to another by adjusting its degree of bending or straightening in response to the driving force provided by the drive unit.
• A positioning unit formed of multiple joints that can be at least partially inserted into the body and is capable of changing its position.
• A drive unit that generates the driving force required for the positioning unit to change its position.

‍

This invention was developed with support from the Ministry of Education, Science and Technology's TOP Campus Construction project.

Underwater Glider Flight Model Design for Optimizing Entry Angles via Cost Function Minimization

This technology is an environment-adaptive flight model design that calibrates static and dynamic parameters using underwater glider flight characteristics and marine environmental data, and calculates optimal flight model parameters based on cost function minimization.

Existing flight models have technical limitations in that they cannot reflect real-time changes in seawater density, hydrodynamic characteristics due to biofouling, or changes in equipment weight, leading to reduced accuracy in flight trajectory prediction.

This technology inputs test flight data and marine environmental data to numerically calculate vertical velocity and entry angles. It then iteratively optimizes flight parameters—such as parasitic drag coefficients, compressibility, and excess buoyancy—to minimize the cost function between actual vertical trajectory data and the model, thereby updating the final flight model.

‍

‍

Key Features:

• Setting basic flight model parameters using underwater glider measurement data, marine environmental data, and flight characteristic data.
• Calculating the vertical velocity of the underwater glider based on the established basic parameters and setting initial conditions for flight parameters.
• Calculating the entry angle of the underwater glider through numerical computation based on initial conditions and deriving vertical velocity from the flight model.
• Calculating a cost function from the derived vertical velocity and determining the flight parameters that result in the minimum value as the final flight parameters.

‍

Device for Optimizing Buoyancy Adjustment through Density Analysis by Depth in Operational Waters

This technology is a simulation-based control system that designs flight models based on underwater glider flight logs and physical configuration data. It collects environmental data (pressure, temperature, salinity) from the operational area to determine the density structure by depth, then calculates the reference density and required weight adjustments for optimal flight efficiency.

Operational inconvenience and inefficiency arise when the underwater glider must be deployed into the actual sea area to verify density structure, adjusted for weight, and redeployed every time the operational area changes.

This technology combines flight data with physical models to calculate drag coefficients, compressibility, and additional buoyancy. By analyzing environmental data (underwater pressure, temperature, salinity) to calculate the reference density suited to the specific sea area, it provides an algorithm for pre-deployment buoyancy optimization and weight adjustment calculation.

‍

‍

Key Features:

• Underwater glider parameter calculation unit that derives flight parameters representing movement using the glider's flight log data
• Operational area analysis unit that collects underwater pressure and temperature data to calculate density by depth in the operational area
• Reference density analysis unit that calculates the weight change required for the underwater glider to achieve the reference density
• Reference density analysis unit that sets the reference density so that the minimum density during buoyancy adjustment matches the minimum density by depth in the operational area

‍

Self-healing multi-drone network system using hello messages and TTL management

This technology is an algorithm for multi-drone ad-hoc networks that manages neighbor tables using the Time-to-Live (TTL) values of hello messages received from nearby drones. In the event of a communication loss, it autonomously restores network connectivity by moving sequentially to the previous location, the last known location of an expired neighbor, and finally the ground control center.

In Flying Ad Hoc Network (FANET) environments, when communication between drones and the controller is lost, simple "Return to Home" methods cause unnecessary movement, leading to high battery consumption and reduced mission efficiency.

This technology monitors the connection status of neighbor drones in real-time using TTL values. When a network disconnection threshold is reached, it restores the network by moving to specific locations—previous location, expired neighbor location, and ground control center—to re-establish hello message reception.

‍

‍

Key Features:

• Multiple drones that extract IP addresses and location information from hello messages received from other drones
• A step of updating location information and TTL values if neighbor information corresponding to the extracted IP address exists, or creating and recording new neighbor information if it does not
• A step of deleting expired neighbor information from the neighbor table while checking the number of remaining active neighbor drones
• A step of executing the network self-healing mode if the number of active neighbor drones falls below a predefined threshold

‍

Modular mobile robot with automatic face-to-face coupling via laser sensor alignment

This technology is a mechanical coupling control algorithm that allows modular robots with multiple coupling surfaces to measure the distance to another robot using laser sensors and reflective surfaces, then adjust their position and alignment for automatic physical coupling.

Conventional marker-based positioning methods are highly dependent on infrastructure, while coupling methods using LEDs and optical sensors involve complex processing, increased component costs, and longer operation times.

This technology uses laser sensors and reflective surfaces placed on the edges of the coupling surface to measure the relative distance and angle between robots in real time, enabling parallel face-to-face alignment. It ensures tight coupling via connectors (such as magnets or hooks) and dynamically reconfigures idle leg functions based on control center commands upon coupling.

‍

‍

Key Features:

• Configuration where first and second modular mobile robots with multiple coupling surfaces move toward a coupling partner
• Configuration for measuring the distance between the robot and its coupling partner using laser sensors and reflective surfaces on the target coupling surface
• First stage: Moving until the measured distances are equal to achieve face-to-face alignment and proximity to the target coupling surfaces
• Second stage: Engaging the coupling mechanism on the target surfaces to securely connect with the partner once the surfaces are in contact

‍

Bogie device for mobile robots based on rocker-bogie mechanism with active power transmission

This technology is a bogie device and mobile robot that optimizes obstacle traversal efficiency by actively lifting the front wheel module through a frame and power transmission mechanism that connects the front and rear wheel modules based on a rocker-bogie mechanism.

Existing rocker-bogie mechanisms rely solely on friction between the wheels and the ground when overcoming obstacles, making active climbing difficult and limiting efficiency due to insufficient upward rotational force at the front wheels.

This technology secures mechanical stability by directly transmitting power from the drive unit to the front and rear wheel power transmission members within a swing-axis-based rocker-bogie structure, incorporating tension rollers and dampers. It can be applied to outdoor patrol robots, rough-terrain exploration, and delivery robots, allowing them to actively overcome obstacles rather than relying on friction.

‍

‍

Key Features:

• A bogie device connected to both the left and right sides of the front area of the frame, and rear wheels connected to the rear area.
• A pair of rocker connecting members coupled to the frame and spaced apart by a predetermined distance on both sides of the connection point.
• A pair of bogie connecting members positioned in the space between the rocker connecting members, into which the swing axis is inserted.
• A configuration of bogie connecting members arranged to rotate relative to the rocker connecting members via the rotation of the swing axis.

Legged robot with synchronized parallelogram linkage for width and height adjustment

This technology is a synchronized legged robot that adjusts its width and height by transmitting the vertical movement of a single actuator to a moving plate, which is connected to multi-jointed legs configured with a parallelogram linkage structure.

Conventional mobile robots often use individual actuators for each wheel-leg, leading to complex control, high manufacturing difficulty, and reduced mobility in narrow spaces or poor storage efficiency due to a fixed width.

This technology features multiple legs pivotally coupled between a support plate and a vertically movable plate, allowing the legs to rotate in synchronization as the moving plate moves up or down. Applicable to indoor delivery robots and narrow-aisle inspection robots, it reduces both manufacturing costs and control complexity by enabling shape transformation with a single actuator.

‍

‍

Key Features:

• A support plate and a moving plate positioned at a predetermined distance above it
• An actuator that raises and lowers the moving plate to adjust its height relative to the support plate
• At least three leg units provided along the perimeter of the moving plate, each equipped with a wheel at the end
• A configuration where the leg units are coupled to rotate in synchronization relative to the movement of the moving plate

Medical Robot

This technology is a medical robot system that focuses radiation on a target point from multiple angles through the relative movement between a bed and a multi-link robotic arm unit that follows a spherical trajectory.

Existing robot-based treatment equipment has faced limitations in targeting accuracy, prolonged treatment times, increased weight due to complex drive mechanisms, and the risk of collisions when using multiple robotic arms.

This technology features a configuration of multiple links and drive members that follow a spherical trajectory centered on the same point, allowing for independent multi-axis rotation control through vertical and horizontal relative movement between the bed and the robotic arm unit. By adjusting the bed's position, it optimizes targeting efficiency for the target point. Applicable to industrial robots and automation systems, it improves the accuracy and efficiency of radiation therapy for cancer treatment, enabling rapid and precise targeting, simplifying control, and reducing treatment or operation times.

‍

‍

Key Features:

• A bed on which the subject is positioned
• A robotic arm unit equipped with an emission member capable of moving within a spherical trajectory where the target point is located at the center of the sphere
• The bed and the robotic arm unit are capable of relative movement in vertical or horizontal directions
• A configuration that moves the subject to the target point by causing relative movement between the bed and the robotic arm unit

‍

This invention was developed with support from the Ministry of Science, ICT and Future Planning for brain mapping-based robot rehabilitation.

Water Rescue System

This technology is a life-saving system consisting of a control unit that monitors the water environment and identifies object locations using multiple stereo cameras, a rescue robot that autonomously navigates to the victim based on real-time paths generated by the control unit, and a base station that supports these operations.

Existing manual rescue equipment and personnel deployment methods face technical limitations in responding quickly to water safety accidents, often failing to secure the golden time required for life-saving.

This technology performs real-time tracking of swimmers' locations and speeds, as well as hazard zone mapping based on stereo camera data. It provides the rescue robot with a navigation path to autonomously reach the victim, enabling rapid rescue. Applicable to logistics, service robots, and autonomous platforms, it enhances the efficiency of water accident management and improves the ability to provide timely rescue in critical situations such as near-drowning.

‍

‍

Key Features:

• A control unit that generates a movement path from the rescue robot's position to the victim's location based on information acquired from the imaging unit and the object/location recognition and tracking unit, and is capable of transmitting and receiving data with the base station.
• An object and location recognition/tracking unit that identifies and tracks the positions of victims and rescue robots in a water environment.
• A rescue unit that rescues the victim when an incident occurs, based on information provided by the control unit.
• The control unit includes an imaging unit equipped with multiple stereo cameras to provide stereo vision.

‍

This invention was developed with support from the Ministry of Education, Science and Technology for the development of sensor-fusion-based public safety threat detection technology.

Greenhouse Layout Optimization Method to Reduce Robot Travel Time Using Genetic Algorithms

This technology utilizes a genetic algorithm to maximize the movement efficiency of robots within a greenhouse. By considering variables such as inter-bed distance, base station locations, and bed cut points, it generates a population of greenhouse layouts and determines the optimal configuration by setting robot travel time as the fitness function.

When greenhouse layouts are fixed, robots often face excessive travel times back to base stations during tasks like pesticide spraying or harvesting, which is exacerbated by limited battery capacity and restricted movement ranges.

This technology employs a genetic algorithm to repeatedly initialize, mutate, crossover, and recombine greenhouse layout populations, ultimately calculating the optimal aisle and bed configuration to minimize the time required for a robot to travel from any point in the greenhouse back to its base station.

‍

‍

Key Features:

• Initializing the greenhouse layout population and calculating the fitness of each layout
• Genetically mutating the greenhouse layout population and recombining them through crossover operations
• Selecting the optimal layout population with high fitness from the recombined layouts
• Determining whether the criteria are met based on the robot's travel time within the selected layout

‍