This technology is an EEG-based wearable robot control device and method that filters out highly correlated components by linking head motion sensor data with independent component analysis (ICA) of EEG signals to remove artifacts caused by head movement, thereby generating training data for classifying gait intention based on pure EEG signals.
Conventional technologies rely on multiple sensors, such as foot pressure sensors, to determine gait intention, which compromises durability. Furthermore, they face technical limitations in accurately reflecting user intent due to noise interference caused by head movement during EEG measurement.
This technology proposes a method of collecting EEG signals per movement unit using head motion sensors and EEG detectors, identifying and removing components highly correlated with head movement through independent component analysis, and then extracting features from the remaining signals. It can be applied to lower-limb rehabilitation and gait assistance robots, accurately detecting a user's gait intention using only EEG signals without the need for additional sensors.
This invention was developed with the support of the Ministry of Science, ICT and Future Planning for the development of vehicle driving and hazard recognition technology through automated brain signal analysis.
This technology controls the navigation of autonomous mobile robots based on path information defined in blocks. It defines surfaces and task markers—including outlines, ways, and speed data—within each path block to perform obstacle detection, collision avoidance, and speed regulation.
Conventional magnetic tape guidance systems incur high reinstallation costs when factory layouts change. Furthermore, they struggle with efficient speed control and flexible task execution because central control systems cannot account for the real-time status of individual robots.
This technology segments movement paths into blocks, allowing users to configure routes and tasks via an interface. The central system aggregates status information from individual robots to calculate speed adjustments or detours in real-time based on estimated arrival times. Applicable to automated transport in factories and warehouses, it enables flexible responses to facility changes and continuous improvements in operational efficiency.
This invention was developed with support from the Ministry of Trade, Industry and Energy for the development of logistics robot systems applicable to wide-area hospital environments.
This technology is a SLAM system for mobile robots that improves localization accuracy by combining environmental feature points and spatial occupancy data extracted from vision sensors with movement data calculated from motion sensors using a probability-based data fusion technique.
When using vision sensors alone, issues such as motion blur during robot movement and decreased accuracy in dynamic environments often lead to cumulative errors in localization and mapping.
This technology proposes a modular SLAM architecture that includes a vision sensor processor, a motion sensor processor, and a third processor that re-estimates the robot's position by fusing this information using probability-based filters such as a Kalman filter. This allows for mutual compensation between sensors, suppressing error accumulation and enabling the creation of precise maps. It can be applied to various indoor autonomous driving applications, including cleaning robots, logistics robots, and service robots, significantly increasing the reliability of localization through sensor fusion.
This technology is an electrically driven SOBW-type surgical device that replaces the conventional mechanical cable-driven system in the end effector of a surgical robot, instead utilizing direct electrical power to perform gripping, pitching, and yawing motions through an integrated motor and gear structure.
In conventional surgical robots, the end effector is driven by cables, which leads to complex connection structures as the number of joints increases or the extension length grows. This results in issues such as backlash caused by long-distance transmission, as well as durability and precision problems due to cable stretching or breakage.
This technology proposes a method of directly placing motors, screw components, and gears inside the end effector to instantly convert electrical energy into mechanical energy. This allows for independent gripping, pitching, and yawing motions without the need for cables, thereby increasing control precision and achieving structural miniaturization. Applicable to the end effectors of laparoscopic surgical robots, it fundamentally resolves the backlash and durability issues associated with cable-driven systems while simultaneously improving the miniaturization and precision of surgical instruments.
This invention was developed with support from the Ministry of Education, Science and Technology for research on the feasibility of developing next-generation laparoscopic surgical tools applying aerospace, electronic, and mechanical engineering.
This technology is a hemiplegia rehabilitation device that measures the movement of a patient's unaffected limb using sensors such as motion capture, electromyography (EMG), and inertial measurement units (IMU). By analyzing this data in real time, it drives the joints of an exoskeleton robot worn on the hemiplegic side, enabling synchronized bilateral limb movement.
Conventional hemiplegia treatment often relies on simple, repetitive motions, resulting in low rehabilitation efficiency. Existing wearable robots function primarily as simple assistive devices, which limits their ability to induce neuroplasticity and makes it difficult to implement systematic rehabilitation that leverages a patient's cognitive illusions.
This technology proposes a system that integrates a control unit—which receives motion data from the unaffected limb to drive the hemiplegic-side robot in real time—with a visual separation device, such as a screen or mirror, that blocks the view of the unaffected limb to make the patient perceive that their hemiplegic side is moving normally. This approach provides effective rehabilitation by inducing neuroplasticity. It can be used for the rehabilitation of stroke patients with hemiplegia, offering superior therapeutic outcomes compared to traditional repetitive training by combining visual illusion with robotic assistance to stimulate neuroplasticity.
This technology is a passive prosthetic arm that uses a spatial four-bar linkage structure to replicate the natural cross-rotation mechanism between the human ulna and radius. It captures the rotational movement of the residual limb and transmits it from the first axis (coupling) to the second and third axes, using a gear set to amplify the rotation and physically extend the range of motion of the wrist.
Existing electric motor-based prosthetics suffer from heavy weight and poor replication of natural human movement. Furthermore, patients with partial forearm amputations often face limitations in pronation and supination due to the loss of the natural ulnar-radial cross-rotation.
This technology features a coupling unit that converts the movement of the residual limb into rotation around a first axis, a base link corresponding to the ulna, and a four-bar linkage fastening structure. By using a rotation amplification unit composed of four types of gears, the input rotation is amplified and transmitted to the wrist, enabling passive movement that closely mimics the forearm rotation of a healthy individual. It can be applied to prosthetics, rehabilitation aids, and wearable devices, extending the range of wrist pronation and supination without the need for motors, thereby reducing both weight and cost.
This invention was developed with support from the Ministry of Science, ICT and Future Planning for the development of biomimetic bionic arm mechanisms.
This technology features a body member and a pair of sliding leg members, integrating the telescopic function of the first leg with the suction capability of the second leg to create a precision gripping and assembly mechanism.
Conventional multi-jointed grippers are structurally unsuitable for installing electronic components in tight spaces, such as inside small mobile device cases, and relying solely on suction methods often results in poor positioning accuracy.
By combining the telescopic length adjustment of the first leg with the vacuum suction of the second leg, this technology performs a multi-stage gripping operation: it picks up and transports objects with a wide span, then reduces the leg spacing and length while maintaining suction to precisely seat components in confined areas. It can be applied to electronic component assembly, precision manufacturing, and small device automation, enhancing assembly accuracy by precisely positioning parts in tight spaces.
This invention was developed with support from the Ministry of Trade, Industry and Energy for the development of process technology, grippers, and assembly techniques for small, precision electronic component assembly in mobile IT products.
This technology is a training device that rehabilitates motor nerves based on EEG by inducing a sense of body ownership, linking a model that mimics a specific part of the human body with a stimulation unit that provides physical feedback.
There have been challenges regarding motor impairment in patients with brain injuries, as well as limitations in existing virtual reality-based rehabilitation methods.
This technology provides a training system that uses joints and actuators to move a model mimicking a specific body part in a real-world environment, while simultaneously applying physical stimulation to both the actual body part and the model to induce neuroplasticity. It can be applied to rehabilitation training, gait assistance, and medical/welfare services, improving motor function in patients with movement disorders by providing feedback and promoting neuroplasticity.
This invention was developed with support from the Ministry of Science, ICT and Future Planning for brain mapping-based robot rehabilitation.
This technology relates to a layer jamming actuator, a drive system for wearable robots that varies stiffness through an enclosure structure capable of sliding and pivoting.
Conventional layer jamming actuator units have limitations in bending and tensile movement, making them difficult to use in various body postures.
This technology prevents interference and enables smooth operation by configuring each enclosure to slide and pivot relative to one another, making it highly effective for wearable robots that require multiple degrees of freedom.
This invention was developed with support from the Ministry of Trade, Industry and Energy’s Engineering Specialized Graduate School Support Program (Plant Engineering), the Ministry of Science and ICT’s Bionic Hand Mechanism Development project, and the Ministry of Science and ICT’s Human-Centered Soft Robotics Technology Research Center.
This technology enables an underwater robot equipped with an ultrasonic camera to generate a 3D point cloud from image data acquired while moving around an object from multiple directions. It then groups the data by movement direction and reconstructs it into a 3D polygon through 2D projection and polygon calculation.
Existing ultrasonic camera methods have limitations in reproducing seafloor objects as realistic 3D shapes because they map 3D information onto a 2D plane.
This technology proposes a method to achieve high-precision 3D modeling by grouping point clouds acquired from multiple angles, performing maximum overlapping polygon calculations, and applying post-processing noise reduction. It can be applied to seafloor exploration, shipwreck searches, and marine structure diagnostics, providing near-realistic shape information even in high-turbidity environments.
This invention was developed with the support of the Smart Underwater Tunnel System Research Center of the Ministry of Science and ICT.
This technology is a robot manipulator and control method that estimates and compensates for nonlinear friction in real-time without linearization. It utilizes only the robot's built-in motor current and encoder-based joint position data, eliminating the need for external force/torque or acceleration sensors by employing an observer equipped with a low-pass filter.
Conventional friction measurement methods often suffer from low cost-efficiency due to the requirement for expensive force/torque sensors, while observer-based methods frequently face issues with reduced estimation accuracy and limited application scope when simplifying nonlinear friction characteristics.
This technology proposes a method that mathematically estimates friction torque within the robot's dynamic equations using an observer with an integrated low-pass filter. It performs calculations while preserving nonlinearity, based on a dynamic model that includes the inertia matrix, Coriolis force, gravity vector, and gear ratio. By compensating for friction without additional sensors, it significantly improves positioning precision, making it ideal for precision assembly and force control tasks.
This invention was developed with support from the Ministry of Science, ICT and Future Planning for human-product haptic simulation technology.
This technology is a differential gear-based variable stiffness robot joint system that uses two independent drive motor inputs to selectively perform joint rotation and stiffness adjustment based on the combination of the motors' rotational directions.
Conventional variable stiffness joints suffer from low drive efficiency and design redundancies, as one motor is dedicated solely to joint actuation while the other is dedicated solely to stiffness control.
This technology proposes a method where a first rotation module converts the motors' same-direction rotational force into joint rotation, while a second rotation module converts opposite-direction rotational force into linear motion to adjust the preload of an elastic member, thereby varying stiffness. Applicable to collaborative and rehabilitation robots, it maximizes both hardware efficiency and output by utilizing both motors.
This technology is a maintenance robot system featuring a modular climbing mechanism that connects the robot's main body to a climbing mobility module via a universal connector, allowing the mobility module to be swapped according to the work environment.
Conventional technology relies on climbing mechanisms fixed to the complex exterior structures of high-rise buildings, leading to cost inefficiencies as it requires developing separate robots or maintaining a large fleet of robots tailored to specific building characteristics.
This technology utilizes a detachable universal connector between the main body and the climbing module, enabling the use of various interchangeable mobility modules such as legged, wheeled, or tracked types, with an auxiliary control unit that automatically recognizes the swapped module. Applicable to exterior cleaning, painting, and facility inspection, it eliminates the need to develop new robots for each building, significantly reducing implementation costs.
This technology provides a connection structure that allows for the detachable coupling of various tools to a general-purpose unmanned aerial vehicle (UAV), creating a multi-rotor UAV system where multiple drones are integrated and controlled to perform specific mechanical tasks.
Previously, there were inefficiencies in having to manufacture dedicated drones for each type of tool, as well as complexities in the control algorithms and system design required for each robot to perform high-difficulty tasks.
This technology introduces a standardized detachable connection and identification structure between the tool and the drone, and proposes a method for controlling multiple UAVs based on an integrated task process received from a control unit. This ensures versatility, allowing a single drone platform to perform a wide range of mechanical tasks. It can be utilized for facility maintenance, construction work, and disaster prevention, significantly increasing the economic efficiency of drone operations by enabling various missions to be performed simply by swapping tools.
This invention was developed with support from the Ministry of Education, Science and Technology’s Convergence Knowledge-Based Creative Mechanical and Aerospace Talent Training Program and the research project on mechanical manipulation control techniques for quadrotor robots.
This technology is a collaborative control and obstacle avoidance method that uses non-holonomic passive decomposition to independently control the formation maintenance of mobile manipulators, object transport, and obstacle avoidance tasks within separated vector spaces.
When multiple mobile manipulators collaborate, tasks such as maintaining manipulator formation, moving objects, and avoiding obstacles often interfere with one another, making precise control difficult. Furthermore, single-path control methods struggle to efficiently handle both obstacle avoidance and task execution simultaneously.
This technology proposes a method that decomposes the state of mobile manipulators into four independent vector spaces: formation changes, object position changes, platform translation and rotation, and movement interference factors, calculating control inputs for each independently. When obstacles are encountered, the system utilizes redundant degrees of freedom to adjust internal configurations and employs potential functions for avoidance, allowing for safe collaboration while maintaining the intended path. Applicable to multi-robot logistics, collaborative transport of large objects, and factory automation, this approach maximizes control efficiency for collaborative robots by achieving task execution and obstacle avoidance simultaneously.
This invention was developed with support from the Ministry of Science, ICT and Future Planning’s research on real-time control and haptic rendering for haptic interaction between multiple remote users, and the Ministry of Education, Science and Technology’s program for fostering creative mechanical and aerospace talent based on convergence knowledge.
