WO2023068440A1 - 로봇 손 시스템 및 로봇 손 제어 방법 - Google Patents
로봇 손 시스템 및 로봇 손 제어 방법 Download PDFInfo
- Publication number
- WO2023068440A1 WO2023068440A1 PCT/KR2021/017622 KR2021017622W WO2023068440A1 WO 2023068440 A1 WO2023068440 A1 WO 2023068440A1 KR 2021017622 W KR2021017622 W KR 2021017622W WO 2023068440 A1 WO2023068440 A1 WO 2023068440A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- robot hand
- pattern
- sensor unit
- unit
- user
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Ceased
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J15/00—Gripping heads and other end effectors
- B25J15/0009—Gripping heads and other end effectors comprising multi-articulated fingers, e.g. resembling a human hand
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J9/00—Program-controlled manipulators
- B25J9/16—Program controls
- B25J9/1612—Program controls characterised by the hand, wrist, grip control
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61F—FILTERS IMPLANTABLE INTO BLOOD VESSELS; PROSTHESES; DEVICES PROVIDING PATENCY TO, OR PREVENTING COLLAPSING OF, TUBULAR STRUCTURES OF THE BODY, e.g. STENTS; ORTHOPAEDIC, NURSING OR CONTRACEPTIVE DEVICES; FOMENTATION; TREATMENT OR PROTECTION OF EYES OR EARS; BANDAGES, DRESSINGS OR ABSORBENT PADS; FIRST-AID KITS
- A61F2/00—Filters implantable into blood vessels; Prostheses, i.e. artificial substitutes or replacements for parts of the body; Appliances for connecting them with the body; Devices providing patency to, or preventing collapsing of, tubular structures of the body, e.g. stents
- A61F2/50—Prostheses not implantable in the body
- A61F2/54—Artificial arms or hands or parts thereof
- A61F2/58—Elbows; Wrists ; Other joints; Hands
- A61F2/583—Hands; Wrist joints
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61F—FILTERS IMPLANTABLE INTO BLOOD VESSELS; PROSTHESES; DEVICES PROVIDING PATENCY TO, OR PREVENTING COLLAPSING OF, TUBULAR STRUCTURES OF THE BODY, e.g. STENTS; ORTHOPAEDIC, NURSING OR CONTRACEPTIVE DEVICES; FOMENTATION; TREATMENT OR PROTECTION OF EYES OR EARS; BANDAGES, DRESSINGS OR ABSORBENT PADS; FIRST-AID KITS
- A61F2/00—Filters implantable into blood vessels; Prostheses, i.e. artificial substitutes or replacements for parts of the body; Appliances for connecting them with the body; Devices providing patency to, or preventing collapsing of, tubular structures of the body, e.g. stents
- A61F2/50—Prostheses not implantable in the body
- A61F2/68—Operating or control means
- A61F2/70—Operating or control means electrical
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J13/00—Controls for manipulators
- B25J13/08—Controls for manipulators by means of sensing devices, e.g. viewing or touching devices
- B25J13/081—Touching devices, e.g. pressure-sensitive
- B25J13/084—Tactile sensors
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J13/00—Controls for manipulators
- B25J13/08—Controls for manipulators by means of sensing devices, e.g. viewing or touching devices
- B25J13/088—Controls for manipulators by means of sensing devices, e.g. viewing or touching devices with position, velocity or acceleration sensors
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J13/00—Controls for manipulators
- B25J13/08—Controls for manipulators by means of sensing devices, e.g. viewing or touching devices
- B25J13/088—Controls for manipulators by means of sensing devices, e.g. viewing or touching devices with position, velocity or acceleration sensors
- B25J13/089—Determining the position of the robot with reference to its environment
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J15/00—Gripping heads and other end effectors
- B25J15/08—Gripping heads and other end effectors having finger members
- B25J15/10—Gripping heads and other end effectors having finger members with three or more finger members
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J9/00—Program-controlled manipulators
- B25J9/16—Program controls
- B25J9/1602—Program controls characterised by the control system, structure, architecture
- B25J9/161—Hardware, e.g. neural networks, fuzzy logic, interfaces, processor
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J9/00—Program-controlled manipulators
- B25J9/16—Program controls
- B25J9/1628—Program controls characterised by the control loop
- B25J9/163—Program controls characterised by the control loop learning, adaptive, model based, rule based expert control
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J9/00—Program-controlled manipulators
- B25J9/16—Program controls
- B25J9/1656—Program controls characterised by programming, planning systems for manipulators
- B25J9/1664—Program controls characterised by programming, planning systems for manipulators characterised by motion, path, trajectory planning
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61F—FILTERS IMPLANTABLE INTO BLOOD VESSELS; PROSTHESES; DEVICES PROVIDING PATENCY TO, OR PREVENTING COLLAPSING OF, TUBULAR STRUCTURES OF THE BODY, e.g. STENTS; ORTHOPAEDIC, NURSING OR CONTRACEPTIVE DEVICES; FOMENTATION; TREATMENT OR PROTECTION OF EYES OR EARS; BANDAGES, DRESSINGS OR ABSORBENT PADS; FIRST-AID KITS
- A61F2/00—Filters implantable into blood vessels; Prostheses, i.e. artificial substitutes or replacements for parts of the body; Appliances for connecting them with the body; Devices providing patency to, or preventing collapsing of, tubular structures of the body, e.g. stents
- A61F2/50—Prostheses not implantable in the body
- A61F2/68—Operating or control means
- A61F2/70—Operating or control means electrical
- A61F2002/704—Operating or control means electrical computer-controlled, e.g. robotic control
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/30—Nc systems
- G05B2219/39—Robotics, robotics to robotics hand
- G05B2219/39527—Workpiece detector, sensor mounted in, near hand, gripper
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/30—Nc systems
- G05B2219/40—Robotics, robotics mapping to robotics vision
- G05B2219/40408—Intention learning
Definitions
- the description below relates to a robotic hand system and a robotic hand control method.
- a prosthetic arm is worn on the amputated arm.
- Such a conventional prosthetic hand was simply worn on the arm by means of a wearer, and had a disadvantage in that it could not provide a separate gripping force.
- a gripping force capable of gripping an object with a prosthetic hand is imparted by an electronic control device.
- An object according to an embodiment is to provide a robot hand system and a robot hand control method for efficiently gripping a target object.
- An object according to an embodiment is to provide a robot hand control system and a robot hand control method having a high success rate of gripping a target object.
- An object according to an embodiment is to provide a robot hand control system and a robot hand control method that can be operated in an intuitive manner to a user.
- a robot hand system in the robot hand system manipulated by a user, includes a robot hand for gripping a target object; a first sensor unit disposed on the robot hand and detecting a posture of the robot hand in real time; a second sensor unit disposed on the robot hand and detecting 3D surface information of the target object appearing based on the robot hand; and a processor for inferring a motion of the robot hand that meets a user's intention based on sensing information of the first sensor unit and the second sensor unit, and operating the robot hand according to the inferred motion;
- the robot hand includes a plurality of frames; and a finger module connected to the plurality of frames and including one or more joints for changing positions of the plurality of frames.
- a point cloud may be formed with respect to the shape of the 3D surface of the target object.
- the processor may determine a relative position of the robot hand in relation to the target object based on the point data.
- the processor may determine an optimal path for operating the inferred motion of the robot hand based on the relative position of the robot hand.
- a machine learning unit that receives learning data based on the sensing information of the first sensor unit and the second sensor unit, in which motions of the robot hand corresponding to the user's intention are known, and machine-learns a pattern of the learning data; and an analyzer configured to analyze a motion of the robot hand that meets a user's intention from measurement data based on sensing information of the first sensor unit and the second sensor unit using a pre-learned machine learning unit.
- the machine learning unit may include: a learning data acquisition unit acquiring learning data according to sensing information of a first sensor unit and a second sensor unit in which motions of the robot hand corresponding to the user's intention are known; a pattern extraction unit receiving each of the learning data from the learning data acquisition unit and extracting a pattern of each data from each of the learning data; and a machine learning model generation unit for generating a machine learning model by labeling motions of the robot hand that meet the user's intention according to the sensing information of the first sensor unit and the second sensor unit corresponding to each extracted pattern.
- the analysis unit may include a pattern matching unit comparing the measurement data and the pattern extracted from the pattern extraction unit to determine whether the pattern matches; and a result output unit that searches for a pattern determined to be matched by the pattern matching unit in the machine learning model and outputs a motion of the robot hand that meets the corresponding user's intention as a result value.
- the processor may feedback-control the robot hand according to the result value of the result output unit.
- the finger module may include a thumb unit performing a thumb function; an index finger portion performing an index finger function; a middle finger portion performing a function of a middle finger; a ring finger portion performing a function of a ring finger; and a little finger part that performs a little finger function.
- the thumb portion includes two frames and two joint portions;
- the index finger, middle finger, ring finger, and little finger may include three frames and three joints, respectively.
- the second sensor unit may be provided in plurality, and each second sensor unit may be disposed inside each of the thumb, index finger, middle finger, ring finger, and little finger portions.
- a method for controlling a robot hand includes providing a robot hand for gripping a target object; a first detection step of detecting a real-time posture of the robot hand; a second detection step of detecting 3D surface information of the target object appearing based on the robot hand; Based on the information detected in the first detection step and the second detection step, inferring a motion of the robot hand that meets the user's intention, and operating the robot hand according to the inferred motion.
- the machine learning step may include a learning data acquisition step of acquiring each learning data according to the information detected in the first detection step and the second detection step in which the operation of the robot hand corresponding to the user's intention is known; a pattern extraction step of extracting a pattern of each data from the learning data acquisition step; And a machine learning model generation step of generating a machine learning model by labeling motions of the robot hand that match the user's intention according to the information detected in the first detection step and the second detection step corresponding to each extracted pattern. can do.
- the analysis step may include a pattern matching step of comparing the measured data and the pattern extracted in the pattern extraction step to determine whether the pattern matches; and a result output step of searching for a pattern determined to be matched in the machine learning model and outputting a motion of the robot hand that meets the user's intention as a result value.
- the processing step may further include a feedback control step of feedback-controlling the robot hand according to the output result value.
- a robot hand system and method for controlling a robot hand may efficiently grip a target object.
- the robot hand system and method for controlling the robot hand may grasp a target object with high probability.
- the robot hand system and method for controlling the robot hand according to an embodiment may be operated in an intuitive manner to a user.
- FIG. 1 is a perspective view of a robotic hand system according to an embodiment.
- FIGS. 2A and 2B are diagrams for explaining an operating principle of a robot hand according to an embodiment.
- 3A and 3B are diagrams for explaining an operation of a robot hand according to an exemplary embodiment.
- FIG. 4 is a block diagram of a robotic hand system according to an embodiment.
- FIG. 5 is a block diagram of a robotic hand system according to an embodiment.
- FIG. 6 is a block diagram of a machine learning unit of a robot hand system according to an embodiment.
- FIG. 7 is a block diagram of an analysis unit of a robot hand system according to an embodiment.
- FIG. 8 is a flowchart of a method for controlling a robot hand according to an embodiment.
- FIG. 9 is a flowchart of a machine learning step for a method for controlling a robot hand according to an embodiment.
- FIG. 10 is a flowchart of an analysis step for a method for controlling a robot hand according to an embodiment.
- FIG. 11 is a flowchart of processing steps for a method for controlling a robot hand according to an embodiment.
- first, second, A, B, (a), and (b) may be used in describing the components of the embodiment. These terms are only used to distinguish the component from other components, and the nature, order, or order of the corresponding component is not limited by the term.
- an element is described as being “connected,” “coupled to,” or “connected” to another element, that element may be directly connected or connected to the other element, but there may be another element between the elements. It should be understood that may be “connected”, “coupled” or “connected”.
- FIG. 1 is a perspective view of a robot hand system 1 according to an embodiment
- FIGS. 2A and 2B are diagrams for explaining the operating principle of the robot hand 11 according to an embodiment
- FIGS. 3A and 3B are It is a diagram for explaining the operation of the robot hand 11 according to an embodiment
- FIG. 4 is a block diagram of the robot hand system 1 according to an embodiment.
- the robot hand system 1 infers the motion of the robot hand 11 that meets the user's intention, and operates the robot hand 11 according to the inferred motion. can make it
- the robot hand system 1 may include a robot hand 11, a first sensor unit 12, a second sensor unit 13, and a processor 14.
- the robot hand 11 may grip the target object O.
- the robot hand 11 may include a finger module 111, and the finger module 111 may include a plurality of frames 1111 and one or more joints 1112.
- the joint part 1112 is connected to the plurality of frames 1111 and may change the position of the plurality of frames 1111 .
- the frame 1111 may be formed in a structure having a longitudinal direction.
- the joint part 1112 includes one or more motors, and the position of the frame 1111 can be changed by rotation of the motor.
- the robot hand 11 may function as a prosthetic hand or industrial tongs. However, this is an example, and the function of the robot hand is not limited thereto.
- the robot hand system 1 can be applied to places of use of various scales.
- the robot hand system 1 may be applied to a terminal clamp of a heavy-duty crane.
- the finger module 111 includes a thumb portion 111a performing a thumb function, an index finger portion 111b performing an index finger function, and a middle finger portion 111c performing a middle finger function. , a ring finger portion 111d performing a ring finger function and a little finger portion 111e performing a little finger function.
- the thumb portion 111a may include two frames 1111 and two joint portions 1112.
- the index finger portion 111b, the middle finger portion 111c, the ring finger portion 111d, and the little finger portion 111e may include three frames 1111 and three joint portions 1112, respectively.
- the thumb portion 111a may be formed to be shorter in length in the longitudinal direction than the index finger portion 111b, the middle finger portion 111c, the ring finger portion 111d, and the little finger portion 111e.
- the first sensor unit 12 is disposed on the robot hand 11 and can detect the posture of the robot hand 11 in real time.
- the first sensor unit 12 may detect the real-time posture of the robot hand 11 by tracking the position of the second sensor unit 13 described later.
- the first sensor unit 12 may be disposed inside the robot hand 11 .
- the inner side means a position where the palm is formed, that is, a position formed on the back surface of the back of the hand.
- the second sensor unit 13 is disposed on the robot hand 11 and can detect 3D surface information of the target object O appearing on the basis of the robot hand 11 .
- the second sensor unit 13 may be provided in plurality.
- Each second sensor part 13 is disposed inside each of the thumb part 111a, the index finger part 111b, the middle finger part 111c, the ring finger part 111d, and the little finger part 111e. It can be.
- the inner side means a position where the palm is formed, that is, a position formed on the back surface of the back of the hand.
- the second sensor unit 13 can detect 3D surface information of the target object O without obstruction of the field of view.
- the processor 14 infers the motion of the robot hand 11 that meets the user's intention based on the sensing information of the first sensor unit 12 and the second sensor unit 13, and the robot hand 11 according to the inferred motion.
- the hand 11 can be operated. Since the processor 14 does not go through an image processing process, the amount of calculation is reduced, and thus the calculation speed can be remarkably improved.
- the processor 14 determines the shape of the 3D surface of the target object O based on the sensing information of the first sensor unit 12 and the second sensor unit 13.
- a point cloud (C) can be formed for
- the processor 14 may determine the relative position of the robot hand 11 in relation to the target object O based on the point data C.
- the processor 14 may determine an optimal path for operating the inferred motion of the robot hand 11 based on the relative position of the robot hand 11 .
- the processor 14 may operate the robot hand 11 to meet the user's intention based on the sensing information of the first sensor unit 12 and the second sensor unit 13. there is. For example, when the target object O comes deeply into the robot hand 11, the processor 14 uses the entire palm of the robot hand 11 while the finger module 111 surrounds the target object O. By doing so, the robot hand 11 can be operated to grip the target object O (eg, FIG. 3A ). In addition, for example, when the target object O is located adjacent to the end of the finger module 111, the processor 14 uses the end of the finger module 111 to grip the target object O with the robot hand ( 11) can be activated (eg Fig. 3b).
- FIG. 5 is a block diagram of a robot hand system 1 according to an embodiment
- FIG. 6 is a block diagram of a machine learning unit 15 of the robot hand system 1 according to an embodiment
- FIG. It is a block diagram of the analysis unit 16 of the robot hand system 1 according to the embodiment.
- the robot hand system (eg, the robot hand system 1 of FIG. 1 ) according to an embodiment uses machine learning to perform a motion A of the robot hand that meets the user's intention. can be diagnosed
- the robot hand system 1 may further include a machine learning unit 15, an analysis unit 16, and a controller 17.
- the robot hand system 1 includes a first sensor unit (eg, first sensor unit 12 in FIG. 1 ) and a second sensor unit (eg, second sensor unit 12 in FIG.
- the pattern P according to the motion A of the robot hand that meets the user's intention may be learned through machine learning from the learning data LD based on the sensing information of the sensor unit 13 .
- the robot hand system 1 compares the measurement data SD based on the sensing information of the first sensor unit 12 and the second sensor unit 13 with a pre-learned pattern P to determine whether or not matching occurs, and matches
- the operation of the robot hand corresponding to the pattern P may be determined as the operation A of the robot hand that meets the user's intention.
- the robot hand system 1 is a robot hand 11 through feedback control that corrects the robot hand (eg, the robot hand 11 of FIG. 1 ) according to the motion A of the robot hand that meets the determined user's intention. ) can be operated to meet the user's intention.
- the machine learning unit 15 receives learning data LD based on the sensing information of the first sensor unit 12 and the second sensor unit 13, in which motions of the robot hand corresponding to the user's intention are known, and the learning data
- the pattern (P) of (LD) can be machine-learned. That is, the machine learning unit 15 may learn the pattern P of data generated according to the motion A of the robot hand that meets the user's intention through supervised learning. Machine learning of the machine learning unit 15 may be performed in advance before the process is performed.
- the machine learning unit 15 may include a learning data acquisition unit 151 , a pattern extraction unit 152 and a machine learning model generation unit 153 .
- the learning data acquisition unit 151 obtains each learning data LD according to the sensing information of the first sensor unit 12 and the second sensor unit 13 where the operation A of the robot hand that meets the user's intention is known. can be acquired.
- the learning data LD may be acquired for each motion A of the robot hand that meets each user's intention. For example, data based on sensing information of the first sensor unit 12 and the second sensor unit 13 having the first operation A1 is acquired as the first learning data LD1, and the second operation A2 Data based on the sensing information of the first sensor unit 12 and the second sensor unit 13 having ) may be acquired as the second learning data LD2.
- the learning data acquisition unit 151 may acquire data measured through a preliminary experiment or data from a database.
- the pattern extraction unit 152 may receive each learning data LD from the learning data acquisition unit 151 .
- a specific pattern P may be formed on the learning data LD sensed in a specific operation due to the specific operation.
- the pattern extractor 152 may extract a pattern P of each data from each received learning data LD.
- the pattern P may be extracted for each learning data LD.
- the pattern extractor 152 may extract the first pattern P1 from the first training data LD1 and the second pattern P2 from the second training data LD2.
- the pattern extractor 152 may extract the pattern P by FFT processing the learning data LD.
- the pattern extraction unit 152 extracts the pattern P through machine learning from a plurality of learning data LD based on the sensing information of the first sensor unit 12 and the second sensor unit 13 performing the same operation. It can be.
- the machine learning model generation unit 153 performs the operation of the robot hand (A) that meets the user's intention according to the sensing information of the first sensor unit 12 and the second sensor unit 13 corresponding to the extracted pattern P ) can be labeled to create a machine learning model (M). That is, for each specific pattern (P), the motion (A) of the robot hand corresponding to the user's intention may be labeled and stored as a machine learning model (M).
- the machine learning model (M) is information that a specific pattern (P) appears in data detected in a specific motion, that is, information that the robot hand 11 operates in a specific motion when a specific pattern (P) appears in the detected data.
- the first machine learning model M1 stores information about the first pattern P1 and the corresponding first motion A1
- the second machine learning model M2 stores the second pattern P2.
- information about the second operation A2 corresponding thereto may be stored.
- the analysis unit 16 uses the pre-learned machine learning unit 15 to determine the user's intention from the measurement data SD based on the sensing information of the first sensor unit 12 and the second sensor unit 13.
- the motion (A) of the robot hand can be analyzed.
- the analysis unit 16 compares the measured data SD with a pre-learned pattern P, and converts the motion of the robot hand corresponding to the matched pattern P to the motion A of the robot hand that meets the user's intention. can be judged by
- the analysis unit 16 may include a pattern matching unit 161 and a result output unit 162 .
- the pattern matching unit 161 may compare the measured data SD and the pattern P extracted by the pattern extraction unit 152 to determine whether the pattern P matches. That is, the pattern matching unit 161 may determine a pattern P matching the measurement data SD.
- the number of patterns P matching the measurement data SD may be plural.
- the pattern matching unit 161 may perform FFT processing on the measurement data SD to determine whether the pattern P matches. Matching the pattern matching unit 161 between the measured data SD and the pre-learned pattern P may be performed through machine learning.
- the result output unit 162 may search for the pattern P determined to be matched by the pattern matching unit 161 in the machine learning model M, and output a corresponding motion of the robot hand as a result value. That is, the result output unit 162 may output the motion of the robot hand corresponding to the pattern P matched to the measurement data SD as the motion A of the robot hand that meets the user's intention.
- a processor may feedback-control the motion of the robot hand according to the result value of the result output unit 162 .
- the processor 14 may correct and control the motion of the robot hand so as to supplement the motion of the robot hand. According to such feedback control, the operation of the robot hand is corrected and controlled in real time according to the sensing information of the first sensor unit 12 and the second sensor unit 13, so that the robot hand 11 meets the user's intention. can make it work.
- the controller 17 may receive input of the motion of the robot hand 11 according to the user's intention.
- the controller 17 may feedback-control the motion of the robot hand 11 according to the motion of the robot hand 11 input through the controller and the motion A of the robot hand that meets the user's intention inferred above.
- FIG. 8 is a flowchart of a robot hand control method 2 according to an embodiment
- FIG. 9 is a flowchart of a machine learning step 25 for the robot hand control method 2 according to an embodiment
- FIG. It is a flowchart of the analysis step 26 of the robot hand control method 2 according to an embodiment
- FIG. 11 is a flowchart of the processing step 24 of the robot hand control method 2 according to an embodiment.
- the robot hand control method 2 diagnoses the motion of the robot hand that meets the user's intention by using machine learning, and inputs the short-channel myoelectric potential signal. It is possible to feedback control the mechanical part according to the user's simple motion intention and the diagnosed necessary motion.
- the robot hand control method (2) includes a robot hand providing step (21), a first detecting step (22), a second detecting step (23), a processing step (24), and a machine learning step (25). and analysis step 26.
- Step 21 of providing a robot hand may be a step of providing a robot hand for gripping a target object.
- the first detection step 22 may be a step of detecting the real-time posture of the robot hand.
- the second detection step 23 may be a step of detecting 3D surface information of a target object appearing based on the robot hand.
- the processing step 24 based on the information detected in the first detection step 22 and the second detection step 23, a motion of the robot hand that meets the user's intention is inferred, and the robot hand motion is inferred according to the inferred motion. It may be a step including an operating step 241 of activating. In one embodiment, the processing step 24 may further include a feedback control step 242 of feedback-controlling the robot hand according to the output result value.
- the machine learning step 25 receives learning data based on the information detected in the first detection step 22 and the second detection step 23 in which motions of the robot hand that match the user's intention are known, and the pattern of the learning data is input. It may be a step of machine learning. In one embodiment, the machine learning step 25 may include learning data acquisition step 251 , pattern extraction step 252 and machine learning model generation step 253 .
- the learning data acquisition step 251 includes each learning data according to the information detected in the first detection step 22 and the second detection step 23 in which the operation of the robot hand that meets the user's intention is known. It may be a step of acquiring.
- the pattern extraction step 252 may be a step of extracting a pattern of each data from the learning data acquisition step 251 .
- the machine learning model generation step 253 labels the motion of the robot hand corresponding to the user's intention according to the information detected in the first detection step 22 and the second detection step 23 corresponding to each extracted pattern. It may be a step of generating a machine learning model by doing so.
- the analysis step 26 may be a step of analyzing the motion of the robot hand that meets the user's intention from the measured data by using the machine learning model learned in advance through the machine learning step 25 .
- the analysis step 26 may include a pattern matching step 261 and a result output step 262 .
- the pattern matching step 261 may be a step of determining whether the pattern matches by comparing the measured data and the pattern extracted in the pattern extraction step 252 .
- the result output step 262 may be a step of searching for a matched pattern in the machine learning model and outputting a motion of the robot hand corresponding to the user's intention as a result value.
Landscapes
- Engineering & Computer Science (AREA)
- Robotics (AREA)
- Mechanical Engineering (AREA)
- Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Orthopedic Medicine & Surgery (AREA)
- General Health & Medical Sciences (AREA)
- Transplantation (AREA)
- Automation & Control Theory (AREA)
- Cardiology (AREA)
- Heart & Thoracic Surgery (AREA)
- Veterinary Medicine (AREA)
- Public Health (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Vascular Medicine (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Physics & Mathematics (AREA)
- Manipulator (AREA)
Abstract
Description
Claims (15)
- 사용자에 의해 조작되는 로봇 손 시스템에 있어서,대상 물체를 파지하기 위한 로봇 손;상기 로봇 손에 배치되고, 상기 로봇 손의 실시간 자세를 검출하는 제1 센서부;상기 로봇 손에 배치되고, 상기 로봇 손을 기준으로 나타나는 상기 대상 물체의 3차원 표면 정보를 검출하는 제2 센서부; 및상기 제1 센서부 및 제2 센서부의 센싱 정보에 기초하여, 사용자의 의도에 부합하는 상기 로봇 손의 동작을 추론하고, 상기 추론된 동작에 따라 상기 로봇 손을 작동시키는 프로세서를 포함하고,상기 로봇 손은,복수개의 프레임; 및 상기 복수개의 프레임에 연결되며 상기 복수개의 프레임의 위치를 변경하기 위한 하나 이상의 관절부를 포함하는, 손가락 모듈을 포함하는, 로봇 손 시스템.
- 제1항에 있어서,상기 프로세서는,상기 제1 센서부 및 제2 센서부의 센싱 정보에 기초하여 상기 대상 물체의 3차원 표면의 형상에 대하여 점자료(point cloud)를 형성하는, 로봇 손 시스템.
- 제2항에 있어서,상기 프로세서는,상기 점자료를 기초로, 상기 대상 물체와의 관계에서의 상기 로봇 손의 상대적 위치를 판단하는, 로봇 손 시스템.
- 제3항에 있어서,상기 프로세서는,상기 로봇 손의 상대적 위치를 기초로, 상기 로봇 손의 추론된 동작의 작동을 위한 최적 경로를 판단하는, 로봇 손 시스템.
- 제1항에 있어서,사용자의 의도에 부합하는 로봇 손의 동작이 알려진 제1 센서부 및 제2 센서부의 센싱 정보에 기초한 학습 데이터를 입력 받고, 상기 학습 데이터의 패턴을 기계학습 하는 기계학습부; 및미리 학습된 기계학습부를 이용하여 제1 센서부 및 제2 센서부의 센싱 정보에 기초한 측정 데이터로부터 사용자의 의도에 부합하는 로봇 손의 동작을 분석하는 분석부를 더 포함하는, 로봇 손 시스템.
- 제5항에 있어서,상기 기계학습부는,사용자의 의도에 부합하는 로봇 손의 동작이 알려진 제1 센서부 및 제2 센서부의 센싱 정보에 따라 각각의 학습 데이터를 취득하는 학습 데이터 취득부;상기 학습 데이터 취득부로부터 각각의 학습 데이터를 전달받고, 각각의 학습 데이터로부터 각각의 데이터의 패턴을 추출하는 패턴 추출부; 및각각의 추출된 패턴에 대응되는 제1 센서부 및 제2 센서부의 센싱 정보에 따른 사용자의 의도에 부합하는 로봇 손의 동작을 라벨링하여 기계학습 모델을 생성하는 기계학습 모델 생성부를 포함하는, 로봇 손 시스템.
- 제6항에 있어서,상기 분석부는,상기 측정 데이터 및 상기 패턴 추출부에서 추출된 패턴을 비교하여 패턴 매칭 여부를 판단하는 패턴 매칭부; 및상기 패턴 매칭부에서 매칭되는 것으로 판단된 패턴을 상기 기계학습 모델에서 탐색하여 대응되는 사용자의 의도에 부합하는 로봇 손의 동작을 결과값으로 출력하는 결과 출력부를 포함하는, 로봇 손 시스템.
- 제7항에 있어서,상기 프로세서는,상기 결과 출력부의 결과값에 따라 상기 로봇 손을 피드백 제어하는, 로봇 손 시스템.
- 제1항에 있어서,상기 손가락 모듈은,엄지 손가락 기능을 수행하는 엄지 손가락부;검지 손가락 기능을 수행하는 검지 손가락부;중지 손가락 기능을 수행하는 중지 손가락부;약지 손가락 기능을 수행하는 약지 손가락부; 및새끼 손가락 기능을 수행하는 새끼 손가락부를 포함하는, 로봇 손 시스템.
- 제9항에 있어서,상기 엄지 손가락부는,두 개의 프레임 및 두 개의 관절부를 포함하며;상기 검지 손가락부, 중지 손가락부, 약지 손가락부 및 새끼 손가락부는,각각 세 개의 프레임 및 세 개의 관절부를 포함하는, 로봇 손 시스템.
- 제10항에 있어서,상기 제2 센서부는 복수개로 구비되며,상기 각각의 제2 센서부는,상기 엄지 손가락부, 검지 손가락부, 중지 손가락부, 약지 손가락부 및 새끼 손가락부 각각의 내측에 배치되는, 로봇 손 시스템.
- 대상 물체를 파지하기 위한 로봇 손을 제공하는, 로봇 손 제공 단계;상기 로봇 손의 실시간 자세를 검출하는 제1 검출 단계;상기 로봇 손을 기준으로 나타나는 상기 대상 물체의 3차원 표면 정보를 검출하는 제2 검출 단계;상기 제1 검출 단계 및 제2 검출 단계에서 검출된 정보에 기초하여, 사용자의 의도에 부합하는 상기 로봇 손의 동작을 추론하고, 상기 추론된 동작에 따라 상기 로봇 손을 작동시키는 작동 단계를 포함하는 프로세싱 단계;사용자의 의도에 부합하는 로봇 손의 동작이 알려진 제1 검출 단계 및 제2 검출 단계에서 검출된 정보에 기초한 학습 데이터를 입력 받고, 상기 학습 데이터의 패턴을 기계학습 하는 기계학습 단계; 및상기 기계학습 단계를 통해 미리 학습된 기계학습 모델을 이용하여, 상기 제1 검출 단계 및 제2 검출 단계에서 측정되는 측정 데이터로부터 사용자의 의도에 부합하는 로봇 손의 동작을 분석하는 분석 단계를 포함하는, 로봇 손 제어 방법.
- 제12항에 있어서,상기 기계학습 단계는,사용자의 의도에 부합하는 로봇 손의 동작이 알려진 제1 검출 단계 및 제2 검출 단계에서 검출된 정보에 따라 각각의 학습 데이터를 취득하는 학습 데이터 취득 단계;상기 학습 데이터 취득 단계로부터 각각의 데이터의 패턴을 추출하는 패턴 추출 단계; 및각각의 추출된 패턴에 대응되는 제1 검출 단계 및 제2 검출 단계에서 검출된 정보에 따른 사용자의 의도에 부합하는 로봇 손의 동작을 라벨링하여 기계학습 모델을 생성하는 기계학습 모델 생성 단계를 포함하는, 로봇 손 제어 방법.
- 제13항에 있어서,상기 분석 단계는,상기 측정 데이터 및 상기 패턴 추출 단계에서 추출된 패턴을 비교하여, 패턴 매칭 여부를 판단하는 패턴 매칭 단계; 및매칭된 것으로 판단된 패턴을 상기 기계학습 모델에서 탐색하여 대응되는 사용자의 의도에 부합하는 로봇 손의 동작을 결과값으로 출력하는 결과 출력 단계를 포함하는, 로봇 손 제어 방법.
- 제14항에 있어서,상기 프로세싱 단계는,출력된 결과값에 따라 상기 로봇 손을 피드백 제어하는 피드백 제어 단계를 더 포함하는, 로봇 손 제어 방법.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US17/622,160 US12544937B2 (en) | 2021-10-20 | 2021-11-26 | Robotic hand system and method for controlling robotic hand |
| EP21961528.3A EP4420843A4 (en) | 2021-10-20 | 2021-11-26 | ROBOT HAND SYSTEM AND ROBOT HAND CONTROL METHOD |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| KR10-2021-0140212 | 2021-10-20 | ||
| KR1020210140212A KR102559105B1 (ko) | 2021-10-20 | 2021-10-20 | 로봇 손 시스템 및 로봇 손 제어 방법 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2023068440A1 true WO2023068440A1 (ko) | 2023-04-27 |
Family
ID=86059224
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/KR2021/017622 Ceased WO2023068440A1 (ko) | 2021-10-20 | 2021-11-26 | 로봇 손 시스템 및 로봇 손 제어 방법 |
Country Status (4)
| Country | Link |
|---|---|
| US (1) | US12544937B2 (ko) |
| EP (1) | EP4420843A4 (ko) |
| KR (1) | KR102559105B1 (ko) |
| WO (1) | WO2023068440A1 (ko) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20220131089A (ko) * | 2021-03-19 | 2022-09-27 | 현대자동차주식회사 | 로봇용 관절 구조체 및 그 구조체를 포함하는 로봇 |
| US12479105B2 (en) * | 2024-12-19 | 2025-11-25 | Digital Global Systems, Inc | Systems and methods of sensor data fusion |
| US12314346B2 (en) | 2024-12-19 | 2025-05-27 | Digital Global Systems, Inc. | Systems and methods of sensor data fusion |
| US12487564B2 (en) | 2024-12-19 | 2025-12-02 | Digital Global Systems, Inc. | Systems and methods of sensor data fusion |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20090118153A (ko) * | 2008-05-13 | 2009-11-18 | 삼성전자주식회사 | 로봇과 로봇 핸드, 로봇 핸드의 제어 방법 |
| JP2019093461A (ja) * | 2017-11-20 | 2019-06-20 | 株式会社安川電機 | 把持システム、学習装置、把持方法、及び、モデルの製造方法 |
| KR20200063353A (ko) * | 2018-11-23 | 2020-06-05 | 강기준 | 카메라와 인공지능을 활용한 다기능 전자 의수 및 그 제어방법 |
| KR20200097572A (ko) * | 2019-02-08 | 2020-08-19 | 한양대학교 산학협력단 | 물체 파지를 위한 훈련 데이터 생성 방법 및 파지 자세 결정 방법 |
| US20210122045A1 (en) * | 2019-10-24 | 2021-04-29 | Nvidia Corporation | In-hand object pose tracking |
Family Cites Families (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9114030B2 (en) | 2007-02-06 | 2015-08-25 | Deka Products Limited Partnership | System for control of a prosthetic device |
| EP2642953B1 (en) | 2010-11-22 | 2016-03-09 | Vanderbilt University | Control system for a grasping device |
| EP2813194B8 (en) * | 2013-06-12 | 2018-07-11 | Otto Bock HealthCare GmbH | Control of limb device |
| US10609359B2 (en) | 2016-06-22 | 2020-03-31 | Intel Corporation | Depth image provision apparatus and method |
| US10773382B2 (en) * | 2017-09-15 | 2020-09-15 | X Development Llc | Machine learning methods and apparatus for robotic manipulation and that utilize multi-task domain adaptation |
| US10682774B2 (en) | 2017-12-12 | 2020-06-16 | X Development Llc | Sensorized robotic gripping device |
| US10792809B2 (en) | 2017-12-12 | 2020-10-06 | X Development Llc | Robot grip detection using non-contact sensors |
| GB2574596A (en) | 2018-06-08 | 2019-12-18 | Epic Inventing Inc | Prosthetic device |
| US20210275807A1 (en) * | 2020-03-06 | 2021-09-09 | Northwell Health, Inc. | System and method for determining user intention from limb or body motion or trajectory to control neuromuscular stimuation or prosthetic device operation |
| PL4302173T3 (pl) * | 2021-03-03 | 2025-09-22 | Guardian Glass, LLC | Systemy i/albo sposoby tworzenia i wykrywania zmian w polach elektrycznych |
| US20240149458A1 (en) * | 2021-03-31 | 2024-05-09 | Honda Motor Co., Ltd. | Robot remote operation control device, robot remote operation control system, robot remote operation control method, and program |
| US12564962B2 (en) * | 2021-03-31 | 2026-03-03 | Honda Motor Co., Ltd. | Robot remote operation control device, robot remote operation control system, robot remote operation control method, and non-transitory computer readable medium |
-
2021
- 2021-10-20 KR KR1020210140212A patent/KR102559105B1/ko active Active
- 2021-11-26 EP EP21961528.3A patent/EP4420843A4/en active Pending
- 2021-11-26 WO PCT/KR2021/017622 patent/WO2023068440A1/ko not_active Ceased
- 2021-11-26 US US17/622,160 patent/US12544937B2/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20090118153A (ko) * | 2008-05-13 | 2009-11-18 | 삼성전자주식회사 | 로봇과 로봇 핸드, 로봇 핸드의 제어 방법 |
| JP2019093461A (ja) * | 2017-11-20 | 2019-06-20 | 株式会社安川電機 | 把持システム、学習装置、把持方法、及び、モデルの製造方法 |
| KR20200063353A (ko) * | 2018-11-23 | 2020-06-05 | 강기준 | 카메라와 인공지능을 활용한 다기능 전자 의수 및 그 제어방법 |
| KR20200097572A (ko) * | 2019-02-08 | 2020-08-19 | 한양대학교 산학협력단 | 물체 파지를 위한 훈련 데이터 생성 방법 및 파지 자세 결정 방법 |
| US20210122045A1 (en) * | 2019-10-24 | 2021-04-29 | Nvidia Corporation | In-hand object pose tracking |
Non-Patent Citations (2)
| Title |
|---|
| HEO SI-HWAN; PARK HYUNG-SOON: "Shared autonomy for the prosthetic hand based on the point cloud information of the object", PROCEEDINGS OF THE 2021 SPRING CONFERENCE OF THE KOREAN SOCIETY OF MECHANICAL ENGINEERS BIOENGINEERING; APRIL 15 - 16, 2021, KOREAN SOCIETY OF MECHANICAL ENGINEERS, BIOENGINEERING DIVISION, KOREA, 16 April 2021 (2021-04-16) - 16 April 2021 (2021-04-16), Korea, pages 75 - 75, XP009546962 * |
| See also references of EP4420843A4 * |
Also Published As
| Publication number | Publication date |
|---|---|
| US20230256618A1 (en) | 2023-08-17 |
| EP4420843A1 (en) | 2024-08-28 |
| KR102559105B1 (ko) | 2023-07-26 |
| EP4420843A4 (en) | 2025-08-27 |
| KR20230056321A (ko) | 2023-04-27 |
| KR102559105B9 (ko) | 2024-04-08 |
| US12544937B2 (en) | 2026-02-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2023068440A1 (ko) | 로봇 손 시스템 및 로봇 손 제어 방법 | |
| WO2019088462A1 (ko) | 혈압 추정 모델 생성 시스템 및 방법과 혈압 추정 시스템 및 방법 | |
| WO2010126321A2 (ko) | 멀티 모달 정보를 이용하는 사용자 의도 추론 장치 및 방법 | |
| WO2011016649A2 (ko) | 얼굴변화 검출 시스템 및 얼굴변화 감지에 따른 지능형 시스템 | |
| WO2022145519A1 (ko) | 딥러닝을 이용한 심전도 시각화 방법 및 장치 | |
| WO2017164478A1 (ko) | 미세 얼굴 다이나믹의 딥 러닝 분석을 통한 미세 표정 인식 방법 및 장치 | |
| WO2016195299A1 (ko) | 탄소섬유강화플라스틱 부품에 대한 광학 검사 방법 | |
| WO2019235828A1 (ko) | 투 페이스 질병 진단 시스템 및 그 방법 | |
| WO2020213755A1 (ko) | 로봇 및 이를 이용한 지도 업데이트 방법 | |
| WO2023025658A3 (de) | Roboterhand eines roboters und verfahren zum trainieren eines roboters sowie ein tragbares sensor- und kraftrückführungselement hierzu | |
| WO2023120777A1 (ko) | 심전도 전역 특징 벡터를 이용한 다중 박동 검출 방법 및 장치 | |
| WO2021086091A1 (ko) | 인공신경망을 이용한 로봇 매니퓰레이터의 충돌을 감지하는 방법 및 시스템 | |
| WO2019208933A1 (ko) | 사용자 인증을 위한 장치 및 방법 | |
| WO2020111327A1 (ko) | 비접촉식 물건속성 인식장치 및 방법 | |
| WO2021137465A1 (ko) | 진공 그리퍼를 이용해 파지 대상물을 파지하여 이송하는 시스템 및 방법 | |
| WO2023163376A1 (ko) | 가상협업 비대면 실시간 원격 실험 시스템 | |
| WO2023177213A1 (ko) | 객체의 색상을 결정하는 방법, 그 장치 및 그 명령을 기록한 기록 매체 | |
| WO2023042991A1 (ko) | 심폐소생술 중 예측 혈압을 제공하는 장치 및 그 방법 | |
| WO2012077909A2 (ko) | 근전도 센서와 자이로 센서를 이용한 지화 인식 방법 및 장치 | |
| WO2019151555A1 (ko) | 물체 검출 방법 및 로봇 시스템 | |
| WO2018194227A1 (ko) | 딥러닝을 이용한 3차원 터치 인식 장치 및 이를 이용한 3차원 터치 인식 방법 | |
| Rathnayake et al. | Development of a real-time Hand gesture recognition system for aid of hearing-impaired communication using flex sensors and machine learning algorithms | |
| WO2018034386A1 (ko) | 생체정보 연동형 스마트보드 시스템 및 그 방법 | |
| WO2023200130A1 (ko) | Rpa를 활용한 장비 및 시설 데이터 수집 시스템 및 방법 | |
| WO2020101121A1 (ko) | 딥러닝 기반의 영상분석 방법, 시스템 및 휴대 단말 |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 21961528 Country of ref document: EP Kind code of ref document: A1 |
|
| WWE | Wipo information: entry into national phase |
Ref document number: 2021961528 Country of ref document: EP |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| ENP | Entry into the national phase |
Ref document number: 2021961528 Country of ref document: EP Effective date: 20240521 |
|
| WWG | Wipo information: grant in national office |
Ref document number: 17622160 Country of ref document: US |