This technology is a camera-robot calibration algorithm that estimates the transformation matrix between the vision sensor's image coordinate system and the robot arm's world coordinate system, improving the precision of coordinate transformation through projection error calculation and updating processes.
Conventional manual calibration methods are time-consuming and prone to errors, as they require repetitive manual tasks to acquire hundreds of feature point pairs.
This technology estimates an initial transformation matrix through primary labeling and performs conditional automated secondary labeling and data association based on projection error, iteratively optimizing the transformation matrix with minimal manual intervention. It can be applied to robot gripping, precision measurement, and automated equipment, thereby improving the accuracy and efficiency of automated labeling systems by implementing a calibration method for precise coordinate estimation and transformation.
This invention was developed with support from the Ministry of Science and ICT for the development of a deep learning-based collaborative robot automated round bar labeling system with worker proximity detection.
This technology features a steering unit containing magnetic materials that respond to external magnetic fields, and a position-tracking unit based on quantum dots that absorb and re-emit short-wave infrared light capable of penetrating the body. These are integrated into the tip of a guidewire to enable non-invasive, real-time tracking and steering.
Conventional vascular intervention procedures rely on X-ray imaging to track guidewire positioning, which exposes both patients and medical staff to radiation and carries risks of side effects from contrast agents, such as shock or heart failure.
This technology incorporates a position-tracking unit at the guidewire tip, consisting of a matrix embedded with magnetic materials and micro/nanoparticles (including quantum dots). This allows for precise steering via external magnetic fields and real-time internal positioning via short-wave infrared illumination. Applicable to industrial robotics and automation systems, this technology eliminates the need for radiation, improves position tracking, removes the risks associated with contrast agents, and enables remote control of the guidewire through complex vascular structures.
This invention was developed with support from the Ministry of Health and Welfare for the development of a microrobotic guidewire system for peripheral vascular intervention.
This technology is a magnetic actuation control system that utilizes an electromagnet array with curved magnetic cores to minimize the occupied footprint and concentrate magnetic force toward a target location.
Conventional radial electromagnet arrangements occupy significant space, making them difficult to install due to interference issues with medical imaging equipment such as C-arms, and limiting the precision of magnetic field formation.
By forming the tips of the magnetic cores into a curved shape, this technology allows electromagnets to be arranged on a flat plane while still being oriented toward the target, thereby increasing layout flexibility and improving space efficiency. It can be applied to industrial robots and automation systems to enhance output magnetic fields and magnetic force, allowing for increased control and precision in the movement of self-propelled robots.
This invention was developed with the support of the Ministry of Science and ICT’s project for the development of magnetic multi-sequential multi-bot-based neural network reconstruction platform technology.
This technology involves bonding conductive particles to the surface of an actuator body—made by twisting non-conductive polymer fibers into a coil structure—using a chemical reduction technique. This enables electrothermal actuation via applied electricity and self-sensing of resistance changes based on length variations.
Existing polymer-based actuators suffer from reduced safety due to high operating temperatures (over 200°C), as well as increased weight, bulk, and compromised flexibility in wearable devices caused by the need for separate feedback sensors and mounting components.
This technology imparts conductivity to a body formed by twisting spandex and polyethylene fibers by chemically reducing silver precursors on its surface. It is designed for low-temperature operation (below 80°C) and monitors contraction deformation in real-time by measuring the actuator's own resistance changes, eliminating the need for external sensors. Applicable to robotic gripping, precision measurement, and automated equipment, it improves operating temperature ranges, reduces the need for additional sensors and components, and enhances the monitoring capabilities of fiber-type actuators.
This invention was developed with support from the Ministry of Science and ICT for the development of an in-vivo closed-loop system using electronic sutures for the treatment of inflammatory bowel disease.
This technology is a capsule structure designed for multi-site biological material sampling. It induces active sample adsorption at target locations by sequentially dissolving a pH-responsive outer protective layer and a temperature-controlled inner protective layer (via a heating layer) to expose a polymer layer. It includes a position and orientation control mechanism based on an external magnetic field using a magnetic body housed within the main unit.
Conventional sampling devices rely solely on passive movement via intestinal peristalsis, and their sampling methods depend exclusively on specific pH environments, making precise selective sampling or multi-site discrete sampling impossible.
This technology implements a structure that allows users to sequentially activate specific sampling modules at desired times and locations by applying inner protective layers with different melting points to each module and independently controlling the heating layer of each module through external signal application. Applicable to surgical robots, interventional systems, and medical automation, it provides a non-invasive method, reduces costs, and improves the ability to collect biological materials from within the digestive tract by enabling access to multiple locations within the gastrointestinal tract.
This invention was developed through support for the development of a micro-carrier-based active precision delivery medical device for multi-departmental knee cartilage regeneration.
This technology features a robot device comprising first and second arm units positioned on an arm base and a connecting unit that links them. Each arm unit performs multi-jointed movements through yaw and pitch drive units and a power transmission unit that connects them.
In conventional SCARA robots, multiple independently driven arms are not interconnected, meaning each arm must grasp objects individually, which limits the weight of the objects that can be handled.
This technology secures structural rigidity by coupling the first and second arm units via a connecting unit. By combining and applying the yaw and pitch driving forces of each arm, it improves the ability to grasp heavy objects and enhances locking force. Applicable to industrial robots and automation systems, it increases clamping and locking power while enabling the handling of significantly heavier loads during yaw and pitch movements.
This invention was developed with support for AI-based diagnostic technology and the development of minimally invasive surgical robots for the precision treatment of multi-site vascular bone diseases.
This technology describes an unpowered gait assistance mechanism that utilizes an elastic body to convert changes in distance between the device and the user, occurring during the user's gait, into a compensatory force.
Existing robotic gait assist devices that use external power sources had problems with temporal and spatial constraints, as well as reduced rehabilitation training effectiveness due to passive joint movements.
This technology, therefore, proposes a method that utilizes an elastic body, a link unit, and an action point conversion unit, all integrated into the weight support section, to convert the user's trunk movement force into gait assistance force and deliver it to the lower limbs in sync with the gait cycle.
This technology was developed through the support of the Pan-Government Medical Device R&D Project Group's research project on the development and usability evaluation of an indoor mobile gait rehabilitation device capable of weight support and lower limb muscle assistance.
This technology describes a mechanism for generating an AI model that leverages Visual Grounding technology to extract object category, position, and attribute information from images. This information is then converted into natural language instructions to plan and control a robot's manipulation trajectory.
Existing robot control methods required operators to manually input object coordinates and task details. This resulted in limitations such as the need for fixed object positions and low operational efficiency when generating commands for multiple objects.
This technology proposes a method for generating a training dataset and subsequently training an AI model. This is achieved using a first framework (GVCCI) which comprises: a visual feature extraction module that recognizes objects and extracts features from images; a module that generates context-appropriate natural language instructions; a model that infers targets and positions via a visual grounding model; and a manipulation module that plans the trajectory of a robot arm.
This technology was developed with support from the Institute of Information & Communications Technology Planning & Evaluation (IITP) through a self-directed AI research project focused on solving novel problems.
This technology extracts individual components of robots and obstacles, then predicts collision distances in parallel through pairwise batch operations. It trains a collision distance prediction model based on geometric feature vectors and relative transformation matrices. The minimum value among the predicted pairwise distances is calculated as the global collision distance, which can then be utilized for real-time motion planning.
Previously, high computational complexity led to performance degradation when calculating minimum distances, a crucial step for conventional motion planning algorithms in high-degree-of-freedom robot systems. Furthermore, data-driven learning methods suffered from low flexibility to environmental changes and frequent retraining requirements, limiting their versatility.
This technology proposes a model that learns by extracting relative transformation values and point cloud-based shape feature vectors between robot components and obstacles. By processing these inputs in batches and performing parallel computations, it enhances operational efficiency and provides flexibility to adapt to environmental changes without needing to retrain for specific shape elements.
This technology was developed with support from the Institute of Information & Communications Technology Planning & Evaluation (IITP) through its goal-oriented AI generation and inference research project.
This technology involves doping Mg, Ti, and Zr into a Na-Ni-Mn-Fe-based layered cathode active material to achieve a composite crystal structure where P2 and O3 phases coexist, thereby mitigating lattice deformation during charging and discharging and improving ion mobility.
Conventional transition metal-based layered cathode materials (Na-Ni-Mn-Fe system) have high discharge capacity, but they suffer from a rapid decrease in capacity retention due to structural instability during repeated charge/discharge cycles.
Accordingly, this technology proposes the design of a cathode active material with a composition of Na a Ni b Mn c Fe d Mg e Ti f Zr gO h (e.g., 0.70≤a≤0.80). Specifically, through Mg, Ti, and Zr doping, it expands the c-axis lattice within the crystal structure, thereby enhancing sodium ion diffusion performance and suppressing irreversible phase transitions, which significantly contributes to securing electrochemical cycle life and reversibility.
This technology was developed with support from the National Research Foundation of Korea's research project on 'Development of 4V-class aqueous lithium-ion batteries through AI-based novel lithium salt discovery'.
This technology converts RGB and depth information from images captured by a mobile robot into embedding data via an encoder module. This data is then mapped with the robot's position information to construct grid-based spatial map data. Subsequently, a decoder module generates rendered images from this map, and by learning the differences from the original captured images through a loss function, optimizes the neural network-based map generation model.
Existing grid-based map generation methods suffer from decreased map accuracy due to the accumulation of robot localization errors. They also require significant memory for storing visual information and have slow data processing speeds, making them difficult to apply in real-world robot operating environments.
This technology introduces a deep neural network encoder-decoder architecture to embed features of captured images into a grid. Through efficient position-information-based data recording and rendering processes, it is an excellent technology that can improve real-time environmental perception and localization accuracy.
This technology was developed with support from the Institute of Information & Communications Technology Planning & Evaluation (IITP) (SW Star Lab) research project 'Robot Learning: Efficient, Safe, and Socially Friendly Machine Learning'.
