This technology is a self-supervised mobile robot that converts 3D point cloud data into grid-based elevation maps, extracts multiple terrain features, and generates an AI model to determine traversability through a self-learning algorithm.
Existing manual labeling methods are costly, simulation data often differs from real-world environments, and simple threshold-based rules struggle to provide precise traversability assessments in complex urban settings.
This technology proposes a method that initializes positive samples from previous driving trajectories and negative samples from grids exceeding thresholds, then iteratively refines the model by reclassifying data based on the classifier's inference probability. This allows the model to improve its accuracy autonomously without human manual labeling. It can be applied to outdoor delivery and patrol robots, providing an economical solution that adapts to new environments without the need for separate data collection.
This technology is a laparoscopic camera control robot and a method for adjusting the camera view that uses AI algorithms to analyze real-time laparoscopic surgical footage. It autonomously controls the optimal camera perspective by recognizing surgical instruments, anatomical structures, and surgical actions.
The quality of laparoscopic camera operation has historically been inconsistent, depending on the skill level and fatigue of the surgical assistant. Furthermore, manual operation methods often lead to interruptions in the surgeon's workflow and a decrease in concentration.
This technology proposes a method that adjusts the camera center based on the position of surgical instruments within the laparoscopic video, corrects screen tilt by analyzing anatomical structures and environmental data, and automatically performs zoom-in and zoom-out functions based on the analysis of surgical actions. Applicable to all types of laparoscopic surgery, it reduces reliance on assistant personnel while simultaneously enhancing the surgeon's focus and the overall quality of the procedure.
This invention was developed with support from the Ministry of Science and ICT for the development of an automated rectal cancer surgery stage recognition system based on deep learning analysis of surgical video data.
This technology is a magnetic field synthesis control device that forms a magnetic field of a specific direction and intensity at a desired location by controlling the current applied to multiple coils. It determines the optimal current command within the rated current limit using a minimum infinity-norm current solution.
Previously, limitations in coil rated current often resulted in reduced synthetic magnetic field strength or unintended deviations in direction.
This technology proposes a method that calculates a first current command using the least squares method and a second current command that minimizes the infinity norm, determining the optimal current command within the rated current limit. This enables stable control that maximizes magnetic field synthesis performance without exceeding coil ratings. It can be utilized as a core control technology to maximize coil performance in fields requiring precise magnetic field control, such as magnetically driven microrobots, precision medical devices, and magnetic levitation systems.
This technology is a soft actuator and soft gripper that operates by injecting fluid into a flexible, zigzag-folded chamber to induce expansion. It achieves both linear deployment and bending motions by physically controlling the deployment angle through deployment limiters positioned between the folds.
Pneumatic soft robots have historically faced issues with large footprints due to internal chamber design and a tendency to sag under their own weight when not in operation.
By applying an origami-inspired structure, this technology minimizes size when not in use and utilizes fixed deployment limiters between the folding surfaces to restrict expansion in specific directions during fluid injection. This allows for the control of complex deployment and bending motions within a single actuator. It is suitable for applications in logistics automation, medical assistive devices, and end-effectors for collaborative robots, and is particularly advantageous for equipment where space efficiency is critical due to its foldable, compact storage design.
This invention was developed with support from the Ministry of Trade, Industry and Energy for the development of collaborative assistive robot arms using foldable hybrid soft robot technology.
This technology is a longitudinally deployable vacuum suction cup that automatically deploys and grips objects by adapting to their position and orientation without the need for separate sensing or control, utilizing a mechanism that retracts the gripper body using vacuum pump negative pressure and expands it through external atmospheric pressure.
Conventional rigid cylindrical grippers cannot grip tilted objects, while standard bellows-type grippers are limited to objects at distances shorter than their initial length, and both require additional equipment to accurately detect the position and angle of objects in unstructured environments.
This technology proposes a system combining a pneumatic control system using a vacuum pump and a three-way valve, a deployable gripper body made of flexible polymer, and an external spring positioned between the body and an internal hose. This allows the gripper to automatically retract and generate gripping force upon contact without sensor feedback. It enables gripping without a separate vision system in environments where object shapes and placements are inconsistent, such as logistics picking, food packaging, and agricultural sorting, significantly reducing the implementation costs of automated equipment.
This invention was developed with support from the Human-Centered Soft Robot Technology Research Center of the Ministry of Science and ICT, and the development of collaborative assistive robot arms using foldable hybrid-actuated soft robot technology from the Ministry of Trade, Industry and Energy.
This technology utilizes two groups of steering wires arranged in helical patterns in opposite directions along the backbone of a soft robot. By alternating the distance of each wire from the backbone's center, it cancels out unintended tension interference caused by wire length changes during backbone bending.
In soft mechanisms, relative displacement between the steering wires and the backbone during bending often leads to unintended tension on the end-effector, resulting in reduced steering precision.
This technology features steering wire groups with opposing helical structures placed on the outer surface or inside of a longitudinally extending backbone. By alternating their positions between the inner and outer sides at each helical period, the system mechanically cancels out the length differences caused by backbone bending. Applicable to surgical soft robots, endoscopes, and inspection robots for confined spaces, it enhances steering precision by neutralizing unintended tension during bending.
This invention was developed with support from the Ministry of Trade, Industry and Energy for the development of robot task control technology capable of grasping, manipulating, and using tools on various objects in daily environments based on multimodal perception.
This technology is a soft robotic gripper composed of a multi-layer elastomer sensor skin and a joint utilizing optical waveguide modes. It employs a multi-layer thin-film sensor skin (bottom, core, and top layers) to sense contact force with objects, and incorporates a flexible joint structure with an embedded core sensor that uses optical waveguide modes to measure the bending angle of the joint.
Conventional SDM-based soft grippers involve complex and time-consuming manufacturing processes and lack an integrated sensing structure capable of simultaneously and precisely measuring both contact force and finger bending angles.
This technology utilizes 3D-printed rigid molds to create soft molds, enabling the integrated manufacturing of elastomer-based phalanges, multi-layer sensor skins, and optical waveguide joints. The sensor skin is formed with a multi-layer structure of varying flexibility to detect contact force, while the joint combines a core sensor and an outer shell to detect bending angles. Applicable to logistics picking, precision assembly, and service robots, this solution enables simultaneous measurement of contact force and bending angles while simplifying production through soft molding.
This invention was developed with support from the Ministry of Trade, Industry and Energy for inflatable soft robotic arm technology for the care of the elderly and patients.
This technology is a wearable knee assist device that combines a link structure mounted on the wearer's lower limb with an actuator. It minimizes resistance during walking by overlapping the links and supports loads with minimal power during standing by utilizing the elasticity and mechanical geometric configuration of the links.
Conventional wearable knee assist devices rely entirely on motor drive, leading to high battery consumption, the need for frequent charging, and reduced operational efficiency due to battery capacity limitations.
By combining an actuator located at the joint with a variable-length link equipped with extension-direction elasticity, this technology eliminates actuator resistance in walking mode and minimizes torque load through the geometric configuration of the links in support mode. It is applicable to muscle strength assistance in industrial settings and gait support for the elderly, ensuring practicality for long-term wear without the burden of frequent charging.
This technology is an interface system for human-robot interaction that acquires electromyography (EMG) signals using a metal-rubber electrode array—composed of elastic material and metal particles—integrated into a body-hugging wearable device. It utilizes ultra-short amplifiers within metal shielding to eliminate noise and controls robotic devices based on measured waveforms and signal propagation path information.
Limitations in human neural potential measurement technology have made it difficult to implement accurate bidirectional human-robot interfaces, particularly resulting in low data precision in environments requiring high-sensitivity measurement, such as rehabilitation therapy and virtual reality.
This technology enhances body contact through elastic metal-rubber electrodes infused with metal particles, suppresses noise with metal shielding, and determines the temporal and spatial propagation paths of EMG waveforms in real time. It can be applied to rehabilitation therapy, prosthetic control, and virtual reality interaction, significantly improving the reliability of human-robot interaction through precise biosignal acquisition.
This technology relates to a spinal assist unit and a human-spine-mimicking kyphosis assistive device, specifically a wearable device designed to correct and support spinal curvature while the user is in an upright position.
Conventional spinal orthotics are rigid, significantly restricting torso movement and causing discomfort in daily life. Furthermore, they are difficult to adapt to individual body types, making long-term wear challenging.
This technology utilizes a semi-active approach, connecting multiple spinal assist units to mimic the human spine and adjusting tension via a wire-driven system, allowing for customized support tailored to the wearer's height, width, and weight.
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 safe 100m sprints in 7 seconds and comfortable 12-hour wear, as well as the development of core technologies for string-twist-based, compact, lightweight, high-performance, and highly durable safe drive modules utilizing string surface reinforcement, variable radius pulleys, and hybrid drive control.
This technology relates to a hip-joint elastic suit and its operating method for physical movement assistance, specifically a non-powered wearable suit that uses elastic elements to assist with hip flexion and extension.
Older adults with reduced muscle mass often experience slower walking speeds and decreased stability, leading to a higher risk of falls. However, existing powered assistive devices are heavy and expensive, making them difficult to wear for daily use.
By connecting the main belt to the thigh-worn components using elastic elements and adjusting the assistive force via length-adjustment members, this technology provides a lightweight and convenient way to improve walking speed and stability.
This invention was developed with support from the Ministry of Science and ICT for "Machine Learning and Extended Reality for High-Speed Mutual Adaptation between Users and Wearable Robots," and the Ministry of Trade, Industry and Energy for the "Development of Human-Augmented Hybrid Robot Suits Capable of Safe 100m Sprints in 7 Seconds and Comfortable 12-Hour Wear."
This technology is a method and device that collects muscle activation data for various gait environments using surface electromyography (sEMG) sensors attached to multiple lower limb muscles, such as the rectus femoris, vastus medialis, and tibialis anterior, and uses this data as input for an artificial neural network to estimate and classify the user's gait environment.
Surface electromyography signals are difficult to classify accurately due to their complex patterns and non-linear characteristics, making it challenging to detect transitions in gait environments early enough to control assistive robots effectively.
This technology proposes a method for estimating gait environments by utilizing electromyography profiles from 11 lower limb muscle sites as inputs for an artificial neural network. It can be applied to exoskeleton gait assistive robots and rehabilitation equipment, enabling rapid recognition of changes in the user's gait environment to provide natural and safe assistance.
This invention was developed through the Ministry of Science and ICT's project on early detection algorithms for gait environment transitions based on biosignals using deep learning techniques and support for the unaffected side.
This technology corrects accumulated localization errors along a robot's path using loop closure. It works by having an underwater robot perform an initial 3D scan of an object at a starting position, traverse multiple locations, and then return to the starting position to perform a second scan of the same object.
In underwater environments, localization errors accumulate as the robot moves, which degrades the consistency of 3D scan data. Previously, it was difficult to ensure accurate localization without relying on expensive, high-precision sensors.
This technology proposes a method to correct both yaw sensor errors and path-based localization data by comparing the first and second scans of the same object. It is applicable to underwater tunnel inspections and marine structure surveys, providing an economical solution for obtaining precise 3D data without the need for expensive navigation equipment.
This invention was developed with support from the Smart Underwater Tunnel System Research Center, funded by the Ministry of Science and ICT.
This technology is a joint positioning device that maintains the position of a multi-joint system by redirecting a unidirectional force from a spring balancer via a tension wire to deliver gravity compensation to linear and rotary joints.
In multi-joint robots, installing individual gravity compensation devices for each joint increases the number of components, adds to the overall mass and volume of the system, and leads to structural complexity and inefficiency.
This technology proposes a method where force applied from a single source is transmitted via a tension wire to the connection points of the linear guide rolling unit and the first and second rotary joints. By winding the wire multiple times around the connection and auxiliary connection parts, the required force is efficiently amplified and transmitted. Applicable to industrial robot arms and medical stands, this system supports the entire multi-joint structure with a single compensation device, achieving both weight reduction and structural simplification.
This invention was developed with support from the Ministry of Trade, Industry and Energy for the development of human-centered smart dual-arm transfer assistance robots.
This technology is a robot system and learning data generation method that determines collisions in real-time by generating training data from the variance between control target values and actual measured values of robot joints during non-collision states, and predicting dynamic normal operating ranges using an AI learning model.
Conventional torque sensor-based collision detection involves high hardware costs, motor current-based methods are prone to false positives due to friction, and existing AI approaches often suffer from reduced robot durability during the collection of actual collision data.
This technology proposes a method that calculates time-series maximum and minimum measured values from normal, non-collision operation data using sliding window and moving average techniques, utilizing them as training data to predict dynamic collision ranges. This enables accurate collision detection without the need for additional sensors. It can be applied to safety certification for collaborative robots and industrial manipulators, replacing expensive torque sensors while ensuring both safety and cost-efficiency.
This invention was developed with support from the Ministry of Trade, Industry and Energy for the development of deep reinforcement learning-based collaborative task technology capable of intelligently responding to unstructured work environments, such as assembly tasks.
