Hand Rehabilitation Training Device

This technology is a rehabilitation device that guides wrist and hand movements by combining multi-axis rotating joints (first and second axes) with a grip resistance component.

There is a lack of systematic grip and wrist rehabilitation training for patients with physical impairments such as hand paralysis or finger curling to aid in muscle and joint recovery.

This technology features a base frame, a wrist training unit, and a grip training resistance component that includes a spring to adjust the load according to the user's training status and measure joint angles. It can be applied to rehabilitation training, gait assistance, and medical/welfare services. By providing a comprehensive system that trains hand and wrist movements and controls load resistance based on the user's clamping force, it enhances the effectiveness of hand rehabilitation training.

‍

‍

Key Features:

• A first grip member disposed between the other sides of a pair of support members, which induces a hand-clenching exercise based on a first load training value while gripped by the user.
• A grip training unit connected to the base frame to rotate about a second axis perpendicular to the first axis.
• A grip training resistance unit provided with a first spring disposed between a fixed part and a moving part to provide the first load training value.
• The grip training unit includes a pair of support members, one side of which is rotatably connected to the base frame.

‍

This invention was developed with support from the Ministry of Trade, Industry and Energy for the development of an ICT-based, customized, game-linked modular rehabilitation and exercise platform for the elderly.

Automatic Detachment Device for Automating Manual Carts via Vertical Gripping Mechanism

This technology provides an automatic detachment mechanism that securely couples with one side of a manual transport device using a vertical gripping structure.

Existing manual transport devices are difficult to retrofit for electrification, and replacing them with automated electric carts often requires disposing of the original equipment.

This technology utilizes a gripping module consisting of a housing, an upper gripper, and a vertically movable lower gripping member driven by a drive shaft to firmly secure manual transport devices and integrate them with automation systems. Applicable to logistics, manufacturing automation, and service robots, it reduces the costs of transitioning to automation without the need to discard existing manual equipment.

‍

‍

Key Features:

• Includes an adjustment module connected to the gripping module to control its vertical height, wherein the gripping module comprises a gripping housing with a predetermined vertical length.
• An automatic detachment device for automating manual transport devices, comprising a lower gripping member positioned below the upper gripping part and connected to the gripping housing to allow for vertical movement.
• An upper gripping part positioned at the front upper section of the gripping housing, extending forward with a predetermined width in the front-to-back direction.
• A gripping module that secures one side of the manual transport device to be coupled.

Unmanned store product stocking robot combining a mobile platform, lifting support, and telescopic robotic arm

This technology is a system that integrates a wheeled mobile platform, a vertical lifting support, a telescopic robotic arm, and a specially designed robotic hand with a finger structure for stocking products in retail stores.

There is a need for automated product stocking in unmanned stores to reduce labor costs, as well as the ability to navigate store spaces effectively to manage inventory.

This technology features a main body with wheels, a telescopic lifting arm, and a robotic hand with rotating and sliding fingers to securely grasp and display products. It can be applied to unmanned stores, logistics automation, and service robotics, increasing efficiency in product stocking and inventory management while reducing labor costs.

‍

‍

Key Features:

• A main body equipped with wheels for mobility, serving as the central structure of the robot
• A lifting support that extends upward from the main body by a predetermined length
• A robotic arm connected to the lifting support, capable of moving vertically to handle products
• A robotic hand connected to the front of the robotic arm, comprising a main hand body, a first finger, and a rotatable second finger

Ankle Strength Assist Device

This technology is a non-powered, passive strength assistance mechanism that aids walking through elastic deformation and restorative force generated during joint flexion, utilizing a link structure positioned on the medial and lateral sides of the user's ankle joint along with leaf-spring-type elastic members mounted at the front and rear.

Existing strength assist devices are often complex and heavy, making them difficult to carry and expensive to manufacture, which hinders commercialization. Furthermore, they pose a high risk of ankle sprains during walking if the user's ankle strength weakens.

This technology features rotatable link members coupled between fixation members that support the calf and foot around the ankle joint. By mounting leaf springs (elastic members) on the sides of these links, the device stores and releases elastic energy based on the joint's flexion angle to assist muscle strength. It can be applied to rehabilitation training, gait assistance, and medical/welfare services, improving the ability to perform walking motions smoothly while enhancing user comfort.

‍

‍

Key Features:

• First and second link members connected between a first fixation member and a second fixation member on the medial and lateral sides of the ankle joint
• First and second elastic members mounted on the first and second link members, oriented toward the front or rear of the ankle joint
• First and second fixation members spaced apart around the ankle joint
• An ankle strength assist device where the first or second link member returns to its original state by the restorative force of the first or second elastic member.

‍

This invention was developed with support from the Ministry of Science and ICT for research into human-augmentation wearable healthcare technology.

Path Planning Method and System for Autonomous Orchard Vehicles Using Trunk Detection

This technology is a path planning mechanism for autonomous vehicles that generates navigation routes within orchards by calculating the ratio of local minima to maxima (LL ratio) and the coordinate ratio (x-y ratio) for each segment in near-infrared camera images, then inputting these into a Bayesian classifier to probabilistically estimate the base of tree trunks.

In orchard environments, irregular ground patterns caused by complex weeds, low-hanging branches, and foliage have historically made machine vision-based tree trunk recognition and accurate positioning difficult.

This technology converts images into binary black-and-white images to separate obstacles into segments, applies a Bayesian probability model to the shape information (LL ratio, x-y ratio) of each segment to detect the base of tree trunks, establishes a center line for the driving path based on the extracted trunk positions via linear regression, and improves the algorithm through feedback from detection results. Applicable to logistics transport, service robots, and autonomous platforms, it enhances the accuracy of trunk detection and improves the stability of navigation path data in orchard environments.

‍

‍

Key Features:

• The base detection step involves acquiring the LL ratio (the ratio of the local minimum to the local maximum of a segment) and the x-y ratio (the ratio of the x and y values of the local minimum).
• The image conversion step involves converting the image into a black-and-white format to distinguish obstacles from the background.
• The step of calculating the probability of a tree trunk based on Bayesian classification of the LL ratio and x-y ratio.
• The step where the autonomous vehicle's image processing unit acquires and converts images of the surrounding environment.

‍

This invention was developed with support from the Ministry of Science and ICT for the development of core technologies for next-generation intelligent systems.

Microrobot and Motion Tracking System for Microrobots

This technology tracks the position and rotational movement of an in-vivo microrobot by irradiating it with near-infrared/short-wave infrared light and detecting the specific wavelengths re-emitted by quantum dots placed on the robot's surface using an external detection device.

Conventional X-ray imaging poses risks of radiation exposure and hardware interference with drive systems, while magnetic resonance imaging (MRI) is difficult to configure for real-time tracking. Furthermore, ultrasound and optical microscopy-based techniques suffer from low resolution, depth limitations, and bio-autofluorescence noise, making real-time precision measurement challenging.

By placing first and second quantum dots at different positions on the microrobot body, this technology determines the robot's rotational state based on the difference in intensity of the emitted light. Utilizing the near-infrared to short-wave infrared spectrum, which offers high biological tissue penetration, it enables a real-time monitoring system free from interference and radiation exposure. This improves the accuracy of measuring microrobot movement for applications in robotic gripping, precision measurement, and automated equipment, without the risks of hardware interference or radiation.

‍

‍

Key Features:

• In-vivo mobile microrobot containing a first quantum dot
• A second quantum dot positioned differently from the first to absorb light emitted by the first quantum dot
• Quantum dots placed on a body equipped with a magnetic unit for movement via external electromagnetic fields
• Motion tracking system comprising a light source, a microrobot, and a detection unit for sensing re-emitted light

‍

This invention was developed with support from the Ministry of Science and ICT for the development of human-robot interaction technology and core components for active exercise rehabilitation.

Device for Estimating the Position of a Moving Object Using Monocular Camera Projection and Circular Band Overlap

This technology calculates distance by comparing the size of an object projected in an image captured by a monocular camera with the pre-set physical dimensions of the actual object. It then generates circular band regions by applying differential error ranges based on the object's position within the image, and estimates the current position of the moving object through the overlapping sections of these regions.

In environments using only a monocular camera, issues with uncertainty in accurate distance estimation and reduced precision in position tracking within complex environments have been persistent challenges.

This technology calculates the distance to each object using the width of the projected object in the image, the camera's focal length, and the actual width of the object. It sets varying error ranges based on the projected object's position to construct circular bands centered on the actual object's coordinates, then identifies the position through the intersecting areas. Applicable to autonomous robots, service robots, and logistics transport platforms, it enhances the accuracy and reliability of position estimation in complex environments using only a single camera.

‍

‍

Key Features:

• A device comprising memory for storing instructions and a processor that executes them to perform operations for estimating the position of a moving object.
• An operation of acquiring an image with a camera mounted on a moving object and obtaining the distance between the actual object corresponding to the projected object and the camera's position.
• An operation of determining different error ranges for distances based on the position of the projected object within the image.
• An operation of generating a circular band-shaped region centered on the coordinates of the actual object and estimating the position of the moving object based on this.

‍

This invention was developed with the support of the Ministry of Science and ICT's AI Convergence Innovation Talent Cultivation program.

Lower-limb exoskeleton analysis device that predicts gait state and terrain using CNN analysis of IMU signals

This technology is an AI-based analysis system that preprocesses IMU sensor data from a lower-limb exoskeleton robot into n-channel images. It analyzes gait states using a CNN-based feature network while simultaneously transmitting feature values from intermediate convolutional blocks to a head network to predict the terrain environment (uphill/downhill/flat).

Conventional technologies require separate training for gait state determination and terrain recognition algorithms, which is time-consuming and inefficient. Furthermore, they face limitations in integrated analysis due to the difficulty of securing large-scale data samples.

This technology constructs input data by converting and stacking IMU measurements into 2D channel images and utilizes a multi-output structure based on a common feature network (convolutional blocks) to perform gait state analysis and terrain classification in parallel within a single model. Applicable to rehabilitation training, gait assistance, and medical/welfare services, it integrates gait state and terrain recognition into one model to improve analysis accuracy.

‍

‍

Key Features:

• A gait analysis device for a lower-limb exoskeleton robot, comprising a head network connected to one of the convolutional blocks, which calculates the terrain environment the user is walking on based on data processed by the connected convolutional block.
• A data receiver that receives measurement data from an IMU (Inertial Measurement Unit) sensor while a user wearing a lower-limb exoskeleton robot equipped with the sensor is walking.
• A data preprocessor that generates input data by converting n features included in the measurement data into an n-channel image.
• A feature network with convolutional blocks for calculating the gait state of the exoskeleton robot from the input data.

‍

This invention was developed with the support of the Ministry of Science and ICT's project for developing AI/big data-based integrated gait control solutions for personalized gait support and evaluation for lower-limb exoskeleton robots.

Shoulder Rehabilitation Robot

This technology is a link-based rehabilitation exercise assistance device that converts the rotational force of a drive shaft into a linkage and crank-connecting rod structure to implement shoulder flexion/extension and abduction/adduction movements.

Existing rehabilitation devices are often limited to specific movements or fail to account for scapular plane motion, making natural shoulder rehabilitation difficult and complicating setup due to the use of multiple drive units.

This technology transmits power from a single motor through a four-bar linkage and crank-connecting rod mechanism to selectively implement flexion/extension and abduction/adduction exercises centered on the scapular plane. It can be applied to rehabilitation training, gait assistance, and medical/welfare services, allowing for selective shoulder flexion/extension and abduction/adduction exercises based on the shoulder plane, as well as the scapular movement known to be necessary prior to shoulder rehabilitation in clinical settings.

‍

‍

Key Features:

• A forearm support unit provided on the slider, where the user's forearm is positioned.
• A drive unit that rotates the drive shaft using externally supplied power.
• A slider configured to move linearly along the longitudinal direction of a guide bar, with its rear end rotatably connected to the front end of a connecting rod, moving linearly along the guide bar through the reciprocating motion of the connecting rod.
• A third link bar whose lower front end is rotatably connected to the lower front end of a second link bar, and whose upper rear end is rotatably connected to the rear side of the drive unit, rotating in conjunction with the second link bar.

End-Effector for Robots

This technology is a paddle-type end-effector designed to lift injured persons or objects from the ground. An elastic element mounted at the end of the paddle member physically deforms upon contact with the object; this deformation toggles a switch to detect the insertion state. Additionally, a hinged bracket provides a compliance function to prevent collisions with the object.

Conventional robotic end-effectors are often too thick, making it difficult to insert them between an injured person and the ground. This poses a risk of secondary injury during insertion, while exposed cables lead to durability issues and a lack of reliability in autonomous robotic rescue operations.

This technology features an insertion-sensing unit composed of an elastic element and a switch at the tip of a thin paddle member, which deforms under external force from the object. It also utilizes a compliance structure that connects the paddle to the robot body via a hinged bracket, allowing for angular adjustments that mitigate impact upon contact. Applicable to logistics picking, service robots, and manufacturing automation, it prevents damage and ensures safe insertion of the paddle, thereby enhancing the safety and effectiveness of rescue operations.

‍

‍

Key Features:

• An end-effector for robots where a switch element turns on when the elastic element is deformed and turns off when the elastic element returns to its original state.
• A sensing unit provided on the paddle member to detect whether the paddle member is fully inserted.
• A paddle member mounted to the robot body, designed to be inserted between the object to be transported and the surface.
• The sensing unit is positioned adjacent to the end opposite to where the paddle member is attached to the robot body, and the sensing unit includes an elastic element that deforms under external force applied by the object being transported.

‍

This invention was developed with support from the Ministry of Trade, Industry and Energy for the development of core technologies for rescue robot end-effectors.

Robot Gripper for Inspection

This technology is a gripper module mounted on the end of a multi-degree-of-freedom robot manipulator. It integrates an air chuck, Remote Center Compliance (RCC), a force-torque sensor, and a laser sensor to perform a real-time impedance control mechanism based on the contact force between components and inspection jigs.

Conventional position-based control methods for manufacturing robots struggle to handle minor jamming or alignment errors between components and inspection jigs, often leading to a reliance on operator skill and reduced accuracy in defect detection.

This technology utilizes a position-based impedance control algorithm that calculates component movement paths in real time. It forms a system that performs precise insertion and position/orientation correction by controlling the RCC through feedback from force-torque and laser sensors. Applicable to logistics, service robots, and autonomous platforms, it improves the accuracy and speed of inserting components into test jigs, while reducing damage and increasing operational efficiency compared to manual methods.

‍

‍

Key Features:

• A laser sensor mounted on a laser sensor bracket attached to one side of the main bracket, which senses the movement distance, speed, and insertion depth of the component into the jig.
• A force-torque (F/T) sensor coupled with the RCC via an RCC bracket, which senses the impact applied when the positioning pin is inserted into the jig.
• A robot gripper for inspection featuring a pair of gripper fingers, each equipped with a pad to enhance gripping force.
• An air chuck with its upper section mounted on one side of the bracket, which grips and releases components using a pair of gripper fingers in response to control signals.

‍

This invention was developed with support from the Ministry of Science and ICT’s AI-based Anti-Drone Active Control Technology Development project.

Grasping Learning System for Irregular Objects Using Offline Reinforcement Learning and Distribution Shift Penalties

This technology utilizes an offline reinforcement learning model to derive grasping poses for irregular objects. It extracts the workspace from offline data collected within the robot's operating environment and applies a penalty to actions with high Q-values that are not present in the valid offline dataset, thereby preventing excessive Q-value predictions outside the offline data distribution and optimizing the grasping success rate.

Real-time online reinforcement learning methods carry a high risk of robot damage during data collection, are time-consuming and costly, and are difficult to implement in real-world field applications due to environmental constraints.

This technology employs an offline reinforcement learning-based model to perform training without real-time interaction. By inferring pixel-wise Q-values and applying a penalty to actions with the maximum Q-value that are not included in the existing dataset (valid offline data), the model is prevented from selecting incorrect optimal actions outside the training data distribution, allowing for the precise derivation of 6-DOF grasping poses. It can be applied to logistics picking, service robots, and manufacturing automation, optimizing grasping success rates without the risk of robot damage during data collection.

‍

‍

Key Features:

• Offline reinforcement learning model training step, where the training unit of the learning server trains an offline reinforcement learning model using valid offline data and infers pixel-wise Q-values
• Penalty application step, where the action with the highest Q-value among those not present in the valid offline data is selected as a representative action, and a penalty is applied only to that representative action
• Maximum Q-value action determination step, where it is verified whether an action is the one with the highest Q-value among those determined to be excluded from the valid offline data
• Valid offline data extraction step, where valid offline data corresponding to the workspace is extracted

‍

This invention was developed with the support of the Artificial Intelligence Convergence Innovation Talent Cultivation program by the Ministry of Science and ICT.

6-DOF Object Grasping Pose Learning System Combining Deep Reinforcement Learning and Gram-Schmidt Orthogonalization

This technology converts spatial information acquired through an input device into point cloud data and applies a deep reinforcement learning model to infer the Q-value (success rate) and rotation vector for each point. Based on the inferred values, it uses Gram-Schmidt orthonormalization to determine the 6-DOF grasping pose.

Conventional supervised learning-based grasping techniques require sophisticated dynamic models, limiting their use for grasping unknown objects without CAD data. Furthermore, the presence of multiple objects often leads to occlusion, which reduces grasping success rates.

This technology uses a deep reinforcement learning model to infer grasping positions and angles directly from point clouds without the need for label generation. In the event of a learning failure, it calibrates the reward function model through inverse reinforcement learning and optimizes the 6-DOF grasping pose using the Gram-Schmidt orthogonalization technique. Applicable to logistics picking, service robots, and manufacturing automation, it significantly increases the grasping success rate for unknown objects even without CAD data.

‍

‍

Key Features:

• Deep reinforcement learning model training phase, which learns from point cloud data and infers the Q-value and rotation vector for each point.
• Work area spatial information extraction phase, where the data generation unit extracts the portion corresponding to the robot device's workspace from the first spatial information to create second spatial information.
• Point cloud data generation phase, which includes a spatial information reception phase where the learning server's data generation unit receives the first spatial information.
• Spatial information acquisition phase, which acquires the first spatial information containing the target object for grasping through an input device.

‍

This invention was developed with the support of the Artificial Intelligence Convergence Innovation Talent Cultivation program by the Ministry of Science and ICT.

Method for Determining the Position and Angle of a Microrobot

This technology is an image processing method that calculates the 3D position and angle of a microrobot by performing thresholding and noise removal on top-view and side-view images of the microrobot collected within a simulation environment, followed by setting a Region of Interest (ROI) and applying edge detection and line detection algorithms.

Challenges include reduced mobility efficiency due to the miniaturization of microrobots, increased control difficulty for real-world human applications, and a lack of precise state recognition technology for pre-testing and simulation.

This technology utilizes camera images to identify the initial position of the microrobot, performs ROI setting and length-based noise filtering, and then extracts the slope and length of line segments through edge and line detection to ultimately derive the microrobot's rotation angles (Yaw, Roll, Pitch) and position. It can be applied to industrial robots and automation systems, improving the control of medical microrobots by providing a method to clearly recognize their position and angle within a simulation environment.

‍

‍

Key Features:

• Includes the step of determining the position and angle of the microrobot in 3D space from an image, where the step of determining the position and angle includes identifying the initial position of the microrobot from the image.
• A step of removing noise that is longer or shorter than the length of the microrobot, based on the length of the microrobot displayed in the image, from the noise generated during the process of isolating the Region of Interest (ROI) from the image.
• A step of performing edge detection at the final position displayed in the image to identify points of discontinuity based on image brightness.
• A step of setting the initial position in the image as a Region of Interest (ROI) for the microrobot.

‍

This invention was developed with support from the Ministry of Trade, Industry and Energy for the development of a microrobotic system for the treatment of chronic total occlusion in myocardial infarction.

Electro-adhesive Pneumatic/Hydraulic Gripper

This technology is a soft robotic gripper mechanism that combines a flexible gripping unit, which expands and contracts via a pneumatic/hydraulic chamber, with an electro-adhesive film that utilizes electrostatic attraction. This increases static friction with objects, allowing for the secure gripping of irregularly shaped items.

Conventional motor-driven robotic hands are limited in their ability to grip irregular objects of various shapes and materials due to their rigid construction, while pneumatic/hydraulic soft robotic grippers often suffer from weak gripping force due to the nature of their flexible materials.

This technology places an electro-adhesive film containing a first electrode in the contact area of the flexible gripping unit to generate electrostatic adhesion. By covering the electrode with a thin film that has lower elongation and higher stiffness than the gripper body, it maintains flexibility while locally enhancing gripping force. Applicable to logistics picking, service robots, and manufacturing automation, it improves both clamping force and durability compared to traditional pneumatic grippers.

‍

‍

Key Features:

• First and second gripping units connected to the main body and driven by the internal pneumatic/hydraulic chamber
• The first gripping unit includes a first electro-adhesive film with a first electrode in the area that contacts the object
• The second gripping unit includes a second electro-adhesive film with a second electrode in the area that contacts the object
• Each electro-adhesive film features a structure where the electrode is encased in a thin film with a lower elongation rate than the gripping unit itself

‍

This invention was developed with support from the Ministry of Trade, Industry and Energy for the development of shape-adaptive electro-adhesive grippers capable of picking irregular multi-objects.