Vision Tracking System Combining Distance Sensor Feedforward and Vision Feedback

This technology is a vision tracking system and method that separates a mobile first body from a second body equipped with a vision sensor. It maintains target tracking performance by compensating for the movement of the first body using distance sensor-based feedforward control combined with feedback signals from the vision sensor.

Previously, vibrations or sudden directional changes during the operation of mobile robots caused targets to move out of the field of view of vision sensors fixed to the same body, or resulted in motion blur, leading to reduced recognition accuracy.

This technology proposes a system that detects the movement and rotation of the first body using distance sensors to drive the second body in the opposite direction via a feedforward control system, while simultaneously integrating feedback signals from the vision sensor itself. This allows for real-time correction of the second body's position and orientation, enabling stable target tracking. Applicable to patrol robots, mobile filming equipment, and logistics robots, it significantly enhances the practicality of robot vision systems by maintaining target tracking even during driving vibrations and sharp turns.

‍

‍

Key Features:

• Distance sensor that detects movement, including the translation and rotation of the first body, to generate signals for feedforward control
• Controller that receives signals from the distance sensor to generate control signals for driving the second body
• Actuator that drives the second body, to which the vision sensor is attached, according to the controller's control signals
• Configuration that generates control signals by utilizing distance sensor signals as feedforward signals and vision sensor signals as feedback signals

‍

This invention was developed with support from the Ministry of Knowledge Economy for the development of u-Robot HRI solutions and core component technologies.

Vision tracking system for target following using feedforward compensation based on drive commands

This technology is a vision tracking system that independently controls a first body responsible for the mobile robot's movement and a second body equipped with a vision sensor. It calculates predicted movement information from the first body's drive commands and uses this to calibrate the orientation of the second body in real time.

In conventional systems, the robot's drive unit and vision sensor are fixed to the same body, causing the target to move out of the field of view or resulting in motion blur during movement, which degrades recognition rates and accuracy.

This technology proposes a method of generating control signals for the second body by combining predicted movement information derived from the first body's drive commands with actual movement data from sensors such as inertial measurement units. By using image data as a feedback signal to measure disturbances, it can actively calibrate the orientation of the second body. It can be applied to mobile surveillance robots, camera drones, and autonomous vehicles, significantly improving image recognition accuracy by maintaining a stable focus on targets even while in motion.

‍

‍

Key Features:

• A first drive unit that receives drive commands to operate the first body and provides them for generating predicted movement information
• A control unit that generates first movement information representing the predicted movement of the first body based on the drive commands
• A configuration that generates control signals to operate the second body using the first movement information
• A second drive unit connected between the first and second bodies that operates the second body according to the control signals

‍

This invention was developed with support from the Ministry of Knowledge Economy for the development of u-Robot HRI solutions and core component technologies.

Convertible upper-limb strength-assist manipulator that unfolds only when using foldable links

This technology is a body-mounted, convertible manipulator structure designed to assist with upper-limb strength. It features multiple foldable links and a four-bar linkage-based hand lift, providing a mechanical interlocking mechanism that allows the device to be unfolded only during heavy-duty tasks and folded and secured to the body when not in use.

Conventional fixed-type manipulators suffer from installation space constraints and reduced mobility, while upper-limb robots integrated with lower-limb exoskeleton robots often cause reduced walking speed and lower drive efficiency due to the added load on the lower-limb structure.

Based on a body-mounted frame, this technology utilizes a variable foldable joint structure incorporating link guide members and lift-locking components. The connecting links ensure that the foldable links and the hand lift are synchronized during deployment and retraction. When not in use, the lift is secured tightly against the body using the opening of the lift-locking component and an elastic support. This design enhances mobility and operational convenience, making it suitable for industrial strength assistance, logistics, and rehabilitation.

‍

‍

Key Features:

• A body-mounted frame and foldable links that can be extended or retracted relative to the body
• A hand lift connected to the foldable links, utilizing a four-bar linkage structure to lift heavy objects
• Connecting links that synchronize the foldable links and the hand lift for simultaneous deployment and retraction
• A lift-locking component that secures the hand lift to the body when the foldable links are in the retracted position

‍

This invention was developed with support from the Ministry of Science, ICT and Future Planning for the development of affordable medical assistance robots through the convergence of remote medical services and robotics technology.

Semi-automated vascular intervention robot for controlling the advancement and rotation of catheters and guidewires

This technology is a semi-automated robotic system that precisely controls the axial advancement and rotational movement of catheters and guidewires for vascular intervention procedures. It features a telescopic structure that supports and guides the catheter, and consists of a catheter rotation unit, a guidewire rotation and feed unit, and a transport unit (rack and pinion).

Existing vascular intervention procedures have faced challenges such as radiation exposure for medical staff, long procedure times due to manual operation, and limited vascular application range and high costs associated with the large outer diameters (4mm or more) of existing robotic systems.

This technology implements a semi-automated system that utilizes existing surgical tools while automating the segments where radiation exposure is most concentrated (catheter and guidewire insertion and rotation). The 4-DOF drive mechanism is designed with a telescopic structure to prevent catheter sagging, and its detachable design ensures ease of sterilization and space efficiency. Applicable to vascular interventions, robotic surgery, and medical automation, it reduces radiation exposure for medical staff while improving procedural precision and efficiency.

‍

‍

Key Features:

• A catheter rotation unit that rotates the catheter along its longitudinal axis while it is inserted.
• A guidewire rotation and feed unit, located on one side of the catheter rotation unit, that advances the inserted guidewire longitudinally and rotates it along its axis.
• A transport unit that moves both the catheter rotation unit and the guidewire rotation and feed unit together along the longitudinal direction of the catheter.
• A telescopic unit, located on the other side of the catheter rotation unit, that supports the catheter and can extend or retract along the catheter's longitudinal direction.

‍

This invention was developed with support from the Ministry of Science, ICT and Future Planning for robotic system technology aimed at reducing radiation exposure and improving procedural accuracy in transarterial chemoembolization for liver cancer.

Gripper Control Device

This technology is a gripper control algorithm that calculates contact and gripping forces through physical modeling—accounting for gravitational acceleration, geometric angles between components, and friction coefficients—based on the 3D spatial orientation of a gripper holding a cylindrical object, thereby deriving the optimal driving force.

Although the force required to grip an object varies depending on its spatial orientation, conventional technologies have suffered from reduced operational efficiency because they either provide gripping force for only specific orientations or lack the capability for intelligent gripping force control across all spatial orientations.

This technology precisely controls gripper output by calculating the first and second contact forces between each component and the object, considering the gripper's pitch and roll, and computing real-time gripping and driving forces using formulas that incorporate the object's mass and geometric shape. Applicable to logistics picking, service robots, and manufacturing automation, it improves the efficiency and accuracy of gripper operation control by precisely calculating clamping and driving forces.

‍

‍

Key Features:

• Calculating the first contact force between the first through fourth components and the cylindrical object held by the gripper when the gripper rotates in the pitch direction
• Calculating the second contact force between the first through fourth components and the cylindrical object held by the gripper when the gripper rotates in the roll direction
• First and second components that are formed at a constant angle relative to the ground and are symmetrical to each other
• A fourth component that is symmetrical to the third component and formed above the second component

‍

This invention was developed with support from the Ministry of Trade, Industry and Energy for the development of end-effector technology for rescue robots.

Mold Air Hole Drilling Robot System

This technology implements a 6-DOF air hole drilling system optimized for 3D irregular mold surfaces by combining a sliding joint (2-DOF) that moves along a curved rail based on a spherical coordinate mechanism with a rotary joint (4-DOF) that controls the drilling tool.

Conventional radial drilling machines are limited to vertical machining, making them unsuitable for irregular curved surfaces. Relying on manual labor leads to reduced efficiency, increased processing time, bottlenecks, and inconsistent production quality.

This technology positions the manipulator on a curved coordinate system using cross-arranged curved rails and sliding joints. It performs automated drilling by driving rotary and linear joints based on control signals derived from drawing analysis and simulation. Each joint is equipped with an electronic brake to maintain high rigidity during drilling. Applicable to industrial robots and automation systems, it improves the reliability of mold air hole machining and production speed by moving the robot manipulator along the curved coordinate system.

‍

‍

Key Features:

• Robot manipulator including a drilling tool and a rotary joint that positions the drilling tool with 4 degrees of freedom
• Rail along which the robot manipulator moves following a curved coordinate system
• Control signal generation configuration that includes drilling point information based on drawing analysis results
• Sliding joint that moves the manipulator with 2 degrees of freedom after verifying the work path via simulation

‍

This invention was developed with support from the Ministry of Science, ICT and Future Planning for research on neural robot technology based on physical and cognitive interaction.

Robot Motion Restriction Control System Combining Motion Sensors and HMD Eye-Tracking Data

This technology is a robot control system that manages robot movement via a user-worn device consisting of motion sensors and an HMD that tracks eye status, adjusting the robot's motion restriction range based on the user's eye state.

When remotely controlling a humanoid robot using only user gestures, safety accidents can occur because the robot continues to mirror the user's movements even when the user is not actively monitoring the robot's situation.

This technology proposes a method that uses the HMD's eye-tracking camera to determine if the user's eyes are closed, thereby controlling the robot's motion restriction range while streaming the robot's camera feed directly to the HMD. Applicable to remote-operated robots and hazardous task automation, it prevents malfunctions when the user is not looking, ensuring both safety and operational stability.

‍

‍

Key Features:

• A robot that mirrors user movements while operating within a defined motion restriction range
• A user-worn device comprising motion sensors to detect user gestures and an HMD to track eye status
• A management server that controls robot movement based on user gestures detected by motion sensors
• A configuration where the HMD's eye-tracking camera determines if the user's eyes are closed to adjust the motion restriction range

Underwater Image Processing Device for Path Retracing Using Checkpoint Image Comparison

This technology is a navigation device for an underwater image processing unit that saves images and locations as checkpoints during free movement, and calculates and corrects relative position errors by comparing real-time captured images with stored images during retracing.

Using dead reckoning in underwater environments leads to cumulative sensor errors, limiting accurate positioning. In particular, there has been a lack of means to correct a robot's position in featureless underwater environments.

This technology proposes a method that uses feature point matching when feature points are detected in an image, and applies a Fourier transform to derive position errors when they are not. It can be applied to underwater exploration robots and marine structure inspection equipment, ensuring reliability in returning accurately to the original path even in homogeneous underwater environments without feature points.

‍

‍

Key Features:

• A drive unit that moves the underwater image processing device and a sensing unit that measures surrounding environmental information
• A configuration that stores images captured during free movement and the locations where those images were taken as checkpoint information
• A configuration that stores images containing feature points in the checkpoint when feature points are detected in the image
• A configuration that calculates and corrects position errors by comparing sensed images with checkpoint images during retracing

Control Technology for Redundant Robot Manipulators Distinguishing Task Reaction Forces from Collisions Using External Force Observers

This technology is a control method that detects the joint torque of a redundant robot manipulator in real-time and compares it with estimates based on a dynamic model to separate normal force-control reaction forces from abnormal external forces, automatically switching the operation mode during abnormal situations.

Existing force-controlled robots based on joint torque sensors struggle to clearly distinguish between normal task reaction forces and abnormal external collision forces, making it impossible to implement efficient control for preventing safety accidents and protecting the robot during collisions.

This technology proposes a method that constructs an external force estimation observer using joint torque sensor data and a Jacobian matrix, filters out task reaction forces, extracts abnormal external force torque, and calculates a collision detection index. It can be applied to collaborative robots and assembly automation equipment, significantly improving safety by accurately identifying collisions while maintaining normal operations.

‍

‍

Key Features:

• A sensing step that acquires detected joint torque information from the joints of a redundant robot manipulator
• An estimation analysis step where estimated joint torque is calculated based on the detected joint torque information and pre-stored estimation data
• A configuration that eliminates normal force-control reaction forces and extracts only abnormal external force torque through a Jacobian matrix and filtering
• A configuration that calculates a collision detection index to set an emergency stop or operation maintenance mode

Human Leg Tracking Technology Based on SJPDAF with Combined Leg-Pair Grouping

This technology enhances target assignment accuracy by first calculating probabilities through primary matching of multiple measured legs extracted from distance sensors with target legs using the SJPDAF technique, followed by secondary posterior probability calculation through leg-pair based grouping.

Existing technologies often treat leg measurements as independent targets or simply group the two closest legs, which leads to tracking errors in crowded environments when target legs are swapped or only a single leg is detected.

This technology introduces a new posterior probability calculation algorithm that considers not only individual elements but also combinations of two legs by adding a leg-pair grouping step to the existing matching process. It can be applied to service robots that follow people, such as guide robots and luggage transport robots, ensuring stable tracking of the target person without losing them even in crowded spaces.

‍

‍

Key Features:

• A step of extracting multiple measured legs corresponding to human legs from distance sensor scan data
• A step of calculating a first posterior probability by applying measured legs and target legs to the SJPDAF technique
• A step of generating grouped target legs by pairing multiple target legs
• A configuration that assigns targets by calculating a second posterior probability between grouped target legs and grouped measured legs

In-pipe inspection robot with an open-body design and impeller for fluid flow maintenance

This technology features a mobile robot capable of navigating fluid environments by utilizing a motor and impeller positioned along the central axis of a cylindrical open-frame structure, allowing for self-propulsion while maintaining fluid flow. A crushing unit at the front of the motor shaft and an impeller at the rear enable the robot to simultaneously break down, collect, and clear debris while in motion.

Existing robots used in narrow pipes or fluid environments often obstruct fluid flow and struggle to efficiently combine self-propelled movement, data collection, and debris removal.

This technology introduces a cylindrical body with open front and rear ends, a propulsion structure that minimizes flow resistance using a motor and impeller, a rotating shaft-linked debris crusher, and a rear debris collection unit. Air bearings installed on the outer wall prevent collisions and maintain stability, allowing the robot to navigate pipe interiors reliably. Applicable to the inspection and cleaning of water mains and piping systems, it significantly reduces maintenance costs by performing movement and debris removal simultaneously without disrupting fluid flow.

‍

‍

Key Features:

• Robot body with open front and rear ends and a motor rotatably mounted along the central axis
• Impeller installed at the rear of the motor shaft to generate self-propulsion through motor rotation
• Power supply and control unit providing operating power and control signals for the motor and impeller
• Multiple sensors for collecting data within fluids or pipes, and a communication unit for data exchange

Underwater robot implementing flapping motion through a combination of intelligent and directional materials

This technology is a flapping-based underwater robot that achieves combined twisting and bending motions within a flexible base material through the physical integration of intelligent materials that respond to external control signals and directional materials that restrict deformation in specific directions.

Existing structures based on intelligent materials are limited to linear or out-of-plane bending, and technical challenges regarding miniaturization and continuous motion have persisted due to complex structural designs and bulky drive components.

This technology proposes a method to induce a difference in twisting angles between the first and second strokes by designing the placement of intelligent materials and the physical orientation of directional materials. This allows for efficient underwater thrust generation without the need for complex joints or multiple motors. It can be applied to underwater exploration, marine monitoring, and small underwater drones, achieving both miniaturization and low power consumption by utilizing material properties for propulsion instead of complex mechanical parts.

‍

‍

Key Features:

• A drive unit comprising intelligent materials that change shape in response to external signals and directional materials that restrict deformation in specific directions
• A robot body connected to the drive unit that converts bending and twisting deformations into underwater propulsion
• A control unit that enables the drive unit to perform a first stroke in one direction and a second stroke in the opposite direction
• A configuration that performs bending and twisting deformations based on the arrangement of intelligent materials and the orientation of directional materials

‍

This invention was developed with support from the Ministry of Education, Science and Technology for biomimetic soft morphing-based technology and the development of design and production technology for multi-scale, multi-deployable collaborative robots.

Mobile Robot SLAM System for Pre-mapping and Expansion via User-Provided Photos

This technology is a SLAM system that generates an initial SLAM map frame by receiving environmental photos and information from a user terminal, and subsequently expands and modifies the map by integrating sensor data collected as the mobile robot navigates.

Conventional SLAM requires robots to explore the entire environment to build a map, which is time-consuming and inefficient, as it often necessitates repeating the entire mapping process to modify or expand parts of an existing map.

This technology proposes a method that uses environmental photos taken from a user terminal to set landmarks and create a basic map framework in advance. When a command to modify or expand the map is received, it performs local updates based on photos of specific areas or moves to the location to integrate real-time data. This reduces mapping time and enables efficient map management. Applicable to home service robots and indoor delivery robots, it significantly reduces initial setup time and enhances user convenience by securing the basic map framework using only photos provided by the user.

‍

‍

Key Features:

• A user terminal that receives environmental information, including photos, and commands from the user and transmits them to the mobile robot.
• A mobile robot that performs mapping by identifying the transmitted environmental information, as well as surrounding environmental information and movement data acquired from a data acquisition device.
• A configuration that pre-builds a portion of the map using environmental information received from the user terminal.
• A configuration that moves to a specific location upon receiving photos of an environment to be modified or expanded, and updates or extends the map accordingly.

Manual lift device with leaf spring elasticity and ratchet locking for assisted lifting

This technology is a manual lift mechanism that utilizes the elastic energy of a leaf spring. It uses a trigger and wire system to control the locking state of a ratchet and pawl, assisting in the lifting and positioning of objects.

Conventional electric lift systems require an external power source, and the use of motors and precision gears leads to high manufacturing costs and installation limitations.

This technology features a dual-ratchet and double-clutch pawl actuator system to mechanically control the winding and unwinding power of a leaf spring. It allows for lifting or securing heavy objects through simple trigger operation without the need for electricity. It can be applied to exoskeleton robots, industrial strength-assist devices, and logistics operations, reducing manufacturing costs and installation constraints by eliminating the need for electrical power.

‍

‍

Key Features:

• A main body rotatably coupled to a frame, with a leaf spring fixed to a lower-limb exoskeleton robot at one end and wound around the body, which is equipped with a hook.
• A pair of ratchets installed on both sides of the main body to restrict its movement as the leaf spring winds or unwinds.
• A ratchet pawl actuator rotatably installed on the main body, which uses a drive protrusion to rotate the pawl and lock or release the ratchet.
• A trigger installed on the outside of the side plate that uses a wire to rotate the ratchet pawl actuator clockwise or counter-clockwise.

5-Bar Linkage-Based Shape-Adaptive Finger Prosthesis Powered by User Movement

This technology is a robotic prosthetic structure based on a 5-bar linkage that attaches to the proximal phalanx of an amputated finger and is powered by the user's own movement.

Conventional robotic prosthetics often stop moving upon contact when grasping objects, leading to unstable grip, or require external actuators, which limits the implementation of prosthetics powered by the body's own movement.

This technology applies a 5-bar linkage mechanism designed to automatically adapt to the shape of an object upon contact by separating the links responsible for flexion/extension and grasping, and incorporating elastic members and angle-limiting elements. It can be applied to prosthetics, rehabilitation aids, and wearable devices, enhancing grasping capability by adapting to various shapes without the need for external actuators.

‍

‍

Key Features:

• A proximal phalanx elastic member provided at the joint between the first and second proximal phalanx links to provide elastic force
• A middle phalanx body connected to the proximal phalanx body, serving the role of the middle phalanx of an amputated finger
• A distal phalanx body connected to the middle phalanx body, serving the role of the distal phalanx of an amputated finger
• A second proximal phalanx link joint-connected to the first proximal phalanx link at the lower side of the proximal phalanx body

‍

This invention was developed with support from the Ministry of Education for research on replaceable bio-finger systems.