INTERFACE FOR THE INTUITIVE MOVEMENT OF A ROBOT AND THE PRECISE REGISTRATION OF ITS TRAJECTORY

Description

Italian Institute of Technology Foundation University of Pisa “Interface for intuitive movement of a robot and the precise recording of its trajectory” 5 ********** The present invention relates to a method and a innovative systems designed to provide a natural and intuitive interface to simplify control of robot movements and improve the 10 accuracy of trajectory recording. Offering an easy-to-use solution, the invention aims to facilitate collaboration between humans and robots, promoting efficiency and precision in different applications. 15 Robots have revolutionized various sectors, offering greater productivity, efficiency and flexibility in performing a wide range of tasks. The ability to control the movements of the robot and accurately record the trajectories is 20 essential for their correct integration in various applications (e.g. welding, painting, distribution of substances, gluing, deburring, grinding, polishing). However, the existing interfaces are often complex and 25 limited intuitiveness and precision. To address these limitations and improve the usability of the robot, an innovative interface is needed and easy to use allowing intuitive control of the robot's movements and the recording of the 30 trajectory. Traditional approaches to robot control are based on complex coding languages ​​and specialized software, which require a wide range of experience in robotics and programming. This limited the accessibility of robots to a group limited number of professionals, hindering adoption widespread and collaboration between humans and robots. 5 Therefore, there is a demand for a more intuitive and easy to use that allows non-users experts in interacting effectively with robots. To meet this need, researchers and engineers have focused on the development of 10 intuitive interfaces for robot movement and trajectory recording. These interfaces aim to simplify the control of robots, allowing users to interact with them in natural and intuitive way. Using technologies 15 advanced features like touch screen, gesture recognition and natural language processing, users can easily specify the location, the desired orientation and trajectory of the robot without the need for complex programming. 20 An effective interface should also provide real-time feedback to users, allowing them to have a clear understanding of the state current state of the robot and its response to their inputs. Visualizations, like virtual representations 25 of the robot and its environment, can help the users to view and fine-tune the desired trajectory before execution. This feedback mechanism increases control of the user and improves the overall experience 30 of the user. Several interfaces have been developed industrial to meet these requirements. A commonly used interface is Teach Pendant described in US 2016 / 0031089 A1, which consists in a portable device with a screen of display and a series of buttons or a joystick for navigation. The Teach Pendant allows the 5 users to directly program and control the robot movements. Operators can move manually the robot into the desired positions and record the trajectory by pressing buttons specific. The Teach Pendant also provides 10 access to various programming options, allowing users to define the points of passage, adjust the speed and configure other parameters. However, using the buttons on the Teach Pendant to move a robot in 3D space 15 has some limitations. 3D control is missing. direct, making it difficult to move the robot in diagonally or along arbitrary paths. Even the simultaneous control of multiple axes is challenging, as it hinders the precise coordination of the 20 position and orientation. The buttons may have limited sensitivity and resolution, influencing speed and accuracy. The feedback tactile is absent, making it difficult to perceive and adjust the force or pressure applied. Use 25 prolonged use of the buttons can cause fatigue and discomfort to the operator due to non-ergonomics optimal. Another method of moving a robot involves direct physical interaction. Some robots are 30 equipped with torque sensors and incorporating controllers that compensate for gravity, allowing users to physically move the robot. This type of interaction allows for direct control and intuitive, as the robot responds to movements of the user's hand. However, it is important to note that not all robots have this specific type of controller implemented. Furthermore, even in the 5 cases where the controller is present, some robots they may have a high level of damping, making it more difficult for the user to handle and manipulate the robot without problems. The presence of Damping can hinder user interaction 10 requiring more effort to move with precision the robot. Therefore, while the manual manipulation provides an intuitive means to control the robot's movements, capabilities and specific characteristics of the robot, including the 15 presence of a controller for compensation of the gravity, as described in Luke, AD, & Panzieri, S. (1993). Learning gravity compensation in robots: Rigid arms, elastic joints, flexible links. International journal of adaptive control and signals 20 processing, 7(5), 417-433 and the damping levels, play a significant role in the ability of the user to manage the robot effectively. Another commonly used interface in industrial environments provides a composite 3D joystick 25 from force / torque sensors as described for example in US 11034022 B2. These sensors are usually located on the robot's end effector, allowing precise control and direct interaction. By applying forces or torques to the sensor, users 30 can manipulate the position and orientation of the robot. Force / torque sensors provide a tactile interface for motion control of the robot, making them suitable for applications that require precise manipulation, such as activities assembly or material handling delicate. However, there are limitations associated to the use of force / torque sensors as 5 interface for controlling the robot's motion. One limitation is the presence of a delay between the application of force and the reaction of the robot to it. This delay, often called latency of the control ring, can affect the 10 system responsiveness and ability to work in real time. The delay is due to factors such as the sensor detection and measurement process force / torque itself, as well as the processing time of the robot control system. This latency 15 may limit the system's ability to perform dynamic activities that require rapid adjustments based on force feedback. Additionally, the use of force / torque sensors as an interface for controlling the movement of the 20 robots present a significant effort in control simultaneous control of both position and orientation. While it is possible to control the location and orientation independently, this separation of control can lead to a loss 25 of intuitive and natural interaction with the robot. Coordinating both aspects of the movement becomes more challenging, potentially hindering the user's ability to perform tasks efficiently efficient and precise. 30 Another noteworthy interface is the pen 3D tracker described in Lin, HI and Lin, YH, 2014. A novel teaching system for industrial robots. Sensors, 14(4), pp.6012-6031. This device laptop is equipped with tracking sensors that capture position and orientation in space 3D. Users can define trajectories you wish by drawing or gesturing in the air, that 5 are then translated into robot commands. The pen 3D tracker offers precise and complex control over robot movements, making it suitable for applications such as painting or welding. However, There are limitations associated with this interface. 10 One limitation is the need for an external sensor, usually an infrared receiver, to monitor the position and orientation of the pen. This dependency introduces further hardware complexity and configuration, with a potential impact on the 15 ease of use and portability of the system. In addition, the performance of the 3D tracker pen can be affected in environments with high reflection levels, which could compromise the accuracy and reliability of the recording 20 of the trajectory. Achieve greater precision with a 3D tracker pen often requires technologies more advanced and expensive tracking, which can limit accessibility and affordability. In conclusion, the development of an interface 25 innovative and intuitive motion control of robots and the recording of trajectories is essential to improve usability and the accessibility of robots in various sectors. This interface should provide control mechanisms 30 intuitive, real-time feedback and the ability to record and edit trajectories. Although the existing interfaces such as Teach Pendants, the gravity compensation controllers, joysticks 3D and 3D tracker pens have their advantages, They also have limitations that may impact on their usability, accuracy and interaction intuitive. Addressing these limitations and developing 5 an advanced interface will allow users to different domains to interact effectively with the robots, promoting collaboration, productivity and further advances in automation and of human-robot interaction. 10 Embodiments of the present invention they concern a learning device for a articulated robot having an end effector controlled by a robot control system. The device includes: 15 a) a kinematic chain having a base and a terminal element; b) a tool coupled to the terminal element of the device; c) a coordinate calculation unit for 20 calculate the position of the tool with respect to the device base; d) a handle fixed to a position of the kinematic chain to induce a movement to the device by a user who grasps such 25 handle; e) a connecting element for connecting in releasable way the device to the robot, such as, for example example, a quick release flange; f) an interface to communicate with the system 30 robot control; where the interface is configured to communicate the trajectory followed by the tool to the robot control system when such tool it is moved by a user acting on the handle of the device in a learning phase for simulate operations that must be performed by the robot when such a tool, or a tool 5 corresponding, is connected to the robot. The unit of calculation of coordinates, or a unit separate control of the device or connected or connectable to the device or to the same system robot control, can then be configured 10 to calculate the tool trajectory with respect to at the base of the robot by applying one or more transformations homogeneous to the positions that form the tool trajectory relative to the base of the device. 15 A can also be advantageously present sensor between the handle and the tool capable of detect the forces / torques applied to the tool. These Detected forces / torques can be read from the unit of coordinate calculation or from the control unit 20 during the learning phase and transmitted to the communication interface to be transmitted to the robot control system to dose the forces applied by the tool in an operational phase. Embodiments also include a robot 25 articulated having a final effector and a system of robot control capable of guiding a tool coupled to the final effector acting on the robot for follow a tool path stored in a memory in a learning phase. The system 30 robot control is configured to derive the desired tool path from a learning device according to the invention when such a learning device is connected to the robot and performs, or has performed, an operation desired tool. The position of the tool relative to the base of the robot can be calculated by the device 5 learning and / or from the robot control system based on position and orientation of the tool relative to the base of the device. The robot can be part of the device learning, that is, it can contribute to movement 10 of the learning device to expand the workspace. In this case the system of robot control can be advantageously configured to control the robot so that follows a transformation that varies over time 15 calculated during trajectory recording in the learning phase. The invention also relates to a method for control the trajectory of an associated tool to the end effector of a robot, the method 20 including: a) record in a first coordinate system the trajectory followed by the tool associated with a learning device connected to the robot when such a device is moved by a user 25 in a training phase to enable to the tool to perform a desired task; b) transform the recorded trajectory into a second coordinate system representing the position and orientation of the tool attached to the 30 robots compared to the robot base; c) guide the robot to follow the trajectory transformed into the second coordinate system as desired path for the tool in one stage operational. The improvements will be subject to the dependent claims. 5 The present invention represents an advance significant in the field of interfacing robots, especially in the area of ​​handling of robots. This innovative solution paves the way to the next step in robot programming, 10 allowing users to move the robot in a more intuitive and easier to use. By simplifying the robot movement control process, the invention eliminates the need for complex coding and programming skills, making it 15 accessible to a wider range of users. This development marks significant milestone in the evolution of robot programming, facilitating greater ease, efficiency and precision in performing tasks. 20 The characteristics of the invention and the advantages that will result from it will be more evident from the description of non-limiting embodiments, illustrated in the attached drawings, in which: Fig. 1 shows a schematic drawing of the 25 kinematic chain that forms a generic N-robot degrees of freedom (Fig. 1a) and of the kinematic chain which forms a generic teaching device for K degrees of freedom (Fig. 1b) to be connected to the robot. Fig. 2 shows a perspective view of a 30 example learning device at 3 degrees of freedom attached to a 6-degree robot freedom. Fig. 3 is an enlarged view of the device of Fig. 2. Fig. 4 is a block diagram showing the robot and the learning device that 5 communicate through a computing unit of the device coordinates / control unit interfaced with the robot control system. Fig. 5 is a variant of the block diagram of Fig. 4 where the calculation unit / unit of 10 control is part of the robot control system. Fig. 6 shows a flowchart of a method for controlling the trajectory of a tool associated with the final effector of a robot according to embodiments of the present invention. 15 The present invention introduces a solution technique that combines a device and a robot to get an extended workspace and control precise for the execution of trajectories and / or profiles of desired force / torque. The integration of the 20 kinematics of the device with the robot's movements improves the usability and versatility of the system, opening up new possibilities for applications that require accurate execution of the trajectory and force interaction. 25 With reference to Fig. 1b, the device according to an embodiment hereof The invention comprises a kinematic chain (active or passive) 102, composed of K degrees of freedom which It could be partially blocked if necessary 30 to simplify its management. The 202 base of the kinematic chain can be fixed at any point of the robot's kinematic chain 101 (shown in Fig. 1a). In this description, for simplicity, the device is considered fixed to the final element 201 of the robot, although such choice is not limiting. The final element 302 of the device is equipped with 5 of a handle that can be gripped ergonomically by the user and of a tool 402 similar to the 401 mounted on the robot. Furthermore, a sensor (not shown) can be incorporated between the handle and the tool 402, capable of capturing 10 the interaction forces / couples applied to the tool 402. An example of a robot could be an ABB robot GoFa (6 degrees of freedom) with a pointed tool. The device can be, for example, a device 15 planar RRR (3 degrees of freedom), also equipped of a 402 tip tool identical to the one mounted on the robot and easily graspable by the user via a handle 602 as shown in Fig. 2. To facilitate device integration 20 with the robot, the base 202 of the device is fixed securely to the robot's end effector (EE) 201 using a 502 quick release flange as shown in Fig. 3. This design allows the device to be easily attached or detached 25 from the robot, ensuring flexibility during use as required. In this specific example, the robot and the device can be advantageously controlled through a web application that can 30 start the registration process of the trajectory and its subsequent replication by the robot. The web application provides an interface intuitive for controlling the device and the robot, allowing the user to interact without problems with the system. In a generic configuration shown in the block diagram of Fig. 4, there is a unit of 5 coordinate calculation 602. This unit could be a dedicated unit of the device or a generic device control unit that acts also from coordinate calculation units for determine the Cartesian coordinates of the tool 10 with respect to the base 202 of the device when such tool is moved by the user in a phase of training. Unit 602 can be interfaced with 501 robot control system or can do part of the 501 robot control system as 15 shown in the configuration of Fig. 5. Information about angles and geometry of the internal joints of the robot and on the position and orientation of the end effector relative to the base of the robot can be supplied to the 602 unit 20 via an entrance to allow you to make calculations in the base coordinate system robot. There may be an element of 612 storage, locally or remotely, 25 where the position of the tool relative to the base of the device can be stored during the learning phase to allow the interface to transfer the overall trajectory followed from the tool to the robot control system 30 when the learning phase is finished. In an alternative configuration, the coordinates points forming the trajectory are sent directly to the 501 robot control system which performs the necessary calculations on the system of coordinates of the robot's base. The purpose of the invention is to generate and record a Cartesian trajectory with the 5 device capturing, preferably, the force / torque profile at the same time. trajectory and force / torque profiles can subsequently be reproduced independently by the robots to perform a specific task. All 10 kinematic properties of the robot are known, so it is direct kinematics is available which allows you to calculate the homogeneous transformation T_Brobot2l(t) where Brobot indicates the base 301 of the el robot any point of the kinematic chain 101, ad 15 example T_Brobot2EE(t) represents the transformation from the robot base 301 to the end effector 201, while T_Brobot2Toolrobot(t) the transformation from robot base 301 to tool tip 401. The transformations involved in a chain 20 kinematics can be obtained using methods known as, for example, the Denavit–Hartenberg method as described in “A simple and systematic approach to assign Denavit-Hartenberg parameters”, IEEE Transactions on Robotics, Vol. 23, No. 3, June 2007 25 is hereby incorporated by reference. To clarify the methodology, let's consider a initial condition in which the robot remains immobile (T_Brobot2EE is a constant value.) At this point, the user can generate a trajectory by grabbing the 30 handle 602. Taking advantage of the kinematics of the device, the homogeneous transformation indicated as T_Bdevice2Tooldevice(t) can be calculated in any time instant t. T_Bdevice2Tooldevice(t) can be expressed with respect to the base of the robot 301 using the following equation: T_Brobot2Tooldevice(t) = T_Brobot2EE * T_EE2Bdevice * 5 T_Bdevice2Tooldevice(t) In this equation, T_Brobot2Tooldevice(t) represents the homogeneous transformation from the base of the robot 301 to the tool 402 of the device, in a 10 given moment t. It is determined by the application sequential of three transformations: T_Brobot2EE, T_EE2Bdevice and T_Bdevice2Tooldevice(t). • T_Brobot2EE represents the transformation homogeneous from the base of the 301 robot to its effector 15 final (EE) 201. Encapsulates angles and geometry of the internal joints of the robot and provides the position and the orientation of the final effector with respect to the robot base. • T_EE2Bdevice represents the transformation 20 homogeneous from the final effector 201 to the base of the device 202. Takes into account any additional transformations and relates the position and orientation of the end effector with respect to the base of the device 202. It can be 25 obtained from CAD or estimated using methods calibrations known in literature as Shiu, Yiu Cheung, and Shaheen Ahmad. "Calibration of wrist- mounted robotic sensors by solving homogeneous transform equations of the form AX= XB." (1987) from 30 be deemed incorporated herein by reference. • T_Bdevice2Tooldevice(t) is the transformation homogeneous from the base of the device 202 to the tool 402 of the device at a specific time t. Describes the position and orientation of the tool 402 of the device compared to its base 202, taking into account any dynamic changes that may occur occur during the trajectory. 5 Multiplying these three transformations in the specified order, we get T_Brobot2Tooldevice(t), which allows you to describe accurately the trajectory of the tool device 402 compared to the robot base 301 in any 10 moment. In the presented scenario, the user can generate a trajectory by manipulating the device. Subsequently, the desired trajectory can be replicated by placing the homogeneous transformation 15 desired T_Brobot2Toolrobot_des(t) equal to T_Brobot2Tooldevice(t). This equivalence guarantees that the robot imitates the same movement as the device. However, it is essential to consider that the handling workspace is restricted 20 from the kinematics of the device. The kinematics imposes limitations on the range of positions reachable and on the orientations that the device can reach. Therefore, the robot's ability to faithfully replicate the generated trajectory is 25 intrinsically limited by kinematic constraints imposed by the device. To extend your workspace, you can change the initial static condition of the robot strategically integrating the robot's kinematics 30 same. This integration constitutes a step fundamental towards achieving a radius expanded operational. As a result, the transformation homogeneous T_Brobot2EE is no longer a constant value, but rather a time-varying parameter that changes dynamically. The robot must track and follow a time-varying reference transformation called T_ref(t), which can be expressed as: T_ref(t) = T_Brobot2EE(t) * T_EE2Bdevice * T_Bdevice2Tooldevice(t) * T_Tooldevice2Bdevice(t_0) * T_Bdevice2EE 10 In this equation: • T_Brobot2EE(t) comes from direct kinematics of the robot and takes into account its configuration changeable at different moments in time. • T_Bdevice2Tooldevice(t) is obtained from the 15 direct kinematics of the device, describing the transformation from its base 202 to the tool 402 of the device at any specific time. • T_Tooldevice2Bdevice(t_0) represents the homogeneous transformation from tool 402 of the 20 device at the base 202 of the device at the beginning of the trajectory recording, indicated as time instant t_0 (typically considered as reference time zero). This transformation captures the initial offset between tool 402 of the 25 device and the device base 202, which is maintained throughout the trajectory. • T_Bdevice2EE represents the transformation homogeneous from the base of the device 202 to the effector final 201. 30 Continuously following T_ref(t) and incorporating the kinematic changes in T_Brobot2EE(t) and T_Bdevice2Tooldevice(t), the robot tries to preserve the relative T_Tooldevice2Bdevice(t_0) offset to the device's tool 402 over time. This collaborative approach allows the user to manipulate the device within a space of expanded work facilitated by combined movements 5 of both the robot and the device, extending made the workspace reachable beyond the intrinsic limitations of the device. The invention also relates to a method for control the trajectory of an associated tool 10 to the final effector of a robot. In the form of realization shown in Fig. 6, the method comprehends: a) record in a first coordinate system the trajectory followed by the tool associated with a 15 learning device connected to the robot when such a device is moved by a user in a training phase to enable to the tool to perform a desired task; b) transform the recorded trajectory into a 20 second coordinate system representing the position and orientation of the tool attached to the robot relative to the robot base; c) guide the robot to follow the trajectory transformed into the second coordinate system as 25 desired tool path in one stage operational. If in phase a) the robot is mobile to expand the workspace, the robot is controlled for follow a transformation that varies over time 30 calculated during trajectory recording in the learning phase considering: a homogeneous transformation that provides the position and orientation of the end effector with respect to the base of the robot; a homogeneous transformation that provides the 5 position and orientation of the tool device relative to the device base; a homogeneous transformation from the tool of the device at the base of the device at the beginning of trajectory recording. 10 In conclusion, the invention provides a method and a device for performing and recording a desired trajectory and, if applicable, the profile force / torque of a robot, which can be reproduced independently at a later time to perform a 15 specific task. The invention focuses on the generation of a trajectory and / or profile force / torque for a robotic system composed of N degrees of freedom and equipped with a tool like reported in Figure 1a. 20 The distinctive features of the invention can be summarized as follows: • The base of the kinematic chain can be fixed at any point of the robot, allowing flexibility in connection. In this case, it is 25 fixed to the last element of the robot, although they are other configurations possible. • Ergonomic design: the final element of the device incorporates a handle that can be ergonomically gripped by the user. The grip 30 is equipped with a tool similar to the one mounted on the robot. • Force / torque sensing: between the handle and the tool has a built-in sensor capable of capture interaction forces and torques applied to the tool. • Expansion of the workspace: for expand the workspace, the dynamic contribution 5 of the robot is incorporated by monitoring a time-varying reference transformation T_ref(t). The robot adjusts its position Cartesian, T_Brobot2EE(t), based on T_ref(t) for preserve the offset over time 10 T_Tooldevice2Bdevice(t_0) related to the tool device. • Collaborative approach: the movements The combined robot and device enable the users to manipulate the device inside 15 an expanded workspace, extending the space of work achievable beyond the intrinsic limits of the device. These advantageous features solve the following problems: 20 • Fixing flexibility: allowing you to fix the base of the kinematic chain in any point of the robot, the device can be fixed in various positions depending on the specific requirements of the task to be performed. This flexibility allows the 25 device to be integrated into different systems robotics without constraints on attachment points. • Versatile configuration: the chain K-degree-of-freedom kinematics offers versatility in the device configuration. Depending on the 30 application-specific needs, the number of degrees of freedom can be adjusted. This allows the customization and optimization of the device functionality to adapt to different tasks and operational requirements. • Lockable kinematic chain: the ability to block the kinematic chain when 5 required offers advantages in simplifying management and device control. By blocking certain degrees of freedom, the device can be stabilized, ensuring precise movements and controlled during generation and reproduction 10 of the trajectory. • Familiarity and acceptance by of the user: the similarity of the grip with the tool mounted on the robot allows a transition and a seamless familiarity from 15 user part. Users accustomed to using directly the robot can easily adapt to the device. Ergonomic design with an interface such a tool facilitates a learning curve smoother and quicker acceptance by 20 of the user, leading to greater efficiency and user satisfaction. • Realistic execution of tasks: the ability to replicate the interaction between the tool and the environment improves the realism of the execution of the 25 activities. The reproduced activity will imitate the behavior of the original activity, including interactions physics and the forces involved. This can be particularly useful in applications where the 30 Precise control of force is essential, as robotic assembly, object manipulation delicate or activities that require compliance of the interaction. • Greater reachable workspace: integrating the kinematic contribution of the robot, the The device's working space is expanded beyond its intrinsic limits. The combined movements of the 5 robots and the device allow users to manipulate the device within a space of broader work, allowing access to positions and orientations that would otherwise be unreachable by device alone. This space 10 extended work offers more flexibility and versatility in performing tasks. Examples of applications in which the invention can be used: Welding: precise control and 15 Recording of robot movements is vital for welding applications, ensuring welds consistent and high quality in the processes of production. Painting: in the automotive and 20 aerospace, the invention's capabilities facilitate precise robot movements for painting flawless and uniform. Dispensing: The invention allows for dispensing controlled and precise handling of fluids and materials, 25 proving particularly valuable in sectors such as the pharmaceutical and electronics manufacturing sectors. Deburring / Grinding / Polishing: The ability of the invention of recording trajectories and forces precision is essential for these processes, 30 improving the quality and consistency of products ended up in sectors such as metalworking, of wood and plastic. The versatility of the invention extends its applications beyond these examples, making it a valuable resource in the automation, industrial robotics, research and development.

Claims

1. A learning device (2) for a articulated robot (1) having an end effector (201) controlled by a robot control system 5 (501), the device comprising: a) a kinematic chain (102) having a base (202) and a terminal element (302); b) a tool (402) coupled to the element device terminal (302); 10 c) a coordinate calculation unit (602) to calculate the tool position (402) with respect to the base of the device (202); d) a handle (602) fixed to a position of the kinematic chain (102) to induce a 15 movement to the device by a user who grab that handle; e) a connecting element (502) for releasably connect the device (2) to the robot (1), such as, for example, a release flange 20 quick; f) an interface to communicate with the system robot control (501); where the interface is configured to communicate the trajectory followed by the tool (402) 25 to the robot control system (501) when such tool is moved by a user acting on the handle of the device (602) in a phase of learning to simulate operations to be performed by the robot when such a tool, or a tool 30 corresponding, is connected to the robot.

2. Device according to claim 1, further comprising a storage element (612) where you can store the position of the tool relative to the base of the device during the learning phase to enable to the interface to transfer the trajectory overall followed by the tool (402) to the system of 5 robot control (501) once the phase is finished of learning.

3. Device according to claim 1 or 2, also including an entrance to receive information on angles and geometries of the 10 internal joints of the robot and on the position and the orientation of the final effector (201) with respect to at the base of the robot, in which the computing unit of the coordinates (602), or a separate control unit of the device or connected or connectable to the 15 device, or to the robot control system (501), is configured to calculate the trajectory of the tool relative to the robot base (301) applying one or more homogeneous transformations to the positions forming the tool trajectory 20 with respect to the base of the device (202).

4. Device according to claim 3, in which the position of the tool in relation to the base of the robot T_Brobot2Tooldevice(t) is calculated using the following equation: 25 T_Brobot2Tooldevice(t) = T_Brobot2EE * T_EE2Bdevice * T_Bdevice2Tooldevice(t) Where: T_Brobot2EE is the homogeneous transformation that provides the position and orientation of the effector 30 final (EE) compared to the base of the robot obtained from the angles and geometry of the internal joints of the robot; T_EE2Bdevice is the homogeneous transformation that provides the position and orientation of the effector final with respect to the base of the device obtained from geometric or estimated calculations using methods of 5 calibration; T_Bdevice2Tooldevice(t) is the transformation homogeneous which provides the position and orientation of the device tool relative to the base of the device. 10 5. Device according to one of the previous claims, further including a sensor between the handle and the tool to detect the forces / torques applied to the tool, such forces / torques being read by the calculation unit 15 of the coordinates or from the control unit during the learning phase and transmitted to the interface communication to be transmitted to the system robot control.

6. An articulated robot (1) having an effector 20 final (201) and a robot control system (501) capable of operating a tool (402) coupled to the final effector (201) acting on the robot to follow a trajectory of the instrument stored in a memory at a time 25 learning, in which the control system of the robot is configured to derive the trajectory desired tool from a device learning according to one or more of the claims previous when such learning device is 30 connected to the robot and performs, or has performed the user's desired operation.

7. Robot according to claim 6, wherein the position of the tool inside with respect to the robot base is calculated by the device learning and / or from the robot control system based on position and orientation of the tool relative to the base of the device 5 (202).

8. Robot according to claim 6 or 7, in which the robot control system (501) is configured to control the robot so that follow a time-varying transform T_ref(t) 10 calculated during trajectory recording in the learning phase according to the following equation: T_ref(t)=T_Brobot2EE(t)*T_EE2Bdevice*T_Bdevice2Toolde vice(t)*T_Tooldevice2Bdevice(t_0)*T_Bdevice2EE 15 Where: T_Brobot2EE(t) is the homogeneous transformation that provides the position and orientation of the effector final (EE) with respect to the base of the robot as derived from the direct kinematics of the robot holding 20 robot configuration modification account at different moments in time; T_Bdevice2Tooldevice(t) is the transformation homogeneous which provides the position and orientation of the device tool relative to the base of the 25 device as obtained from direct kinematics of the device; T_Tooldevice2Bdevice(t_0) is the transformation homogeneous from the device tool to the base of the device at the beginning of recording 30 trajectory (t_0); T_Bdevice2EE is the homogeneous transformation from base of the device to the final effector.

9. Method for controlling the trajectory of a tool associated with the end effector of a robot, the method including: a) record in a first coordinate system 5 the trajectory followed by the tool associated with a learning device connected to the robot when such a device is moved by a user in a training phase to enable to the tool to perform a desired task; 10 b) transform the recorded trajectory into a second coordinate system representing the position and orientation of the tool attached to the robot relative to the robot base; c) guide the robot to follow the trajectory 15 transformed into the second coordinate system as desired path for the tool in one stage operational.

10. The method of claim 9, wherein, if in phase a) the robot is mobile to expand the 20 working space, the robot is controlled to follow a calculated time-varying transformation during the trajectory recording in the phase of learning considering: a homogeneous transformation that provides the 25 position and orientation of the end effector with respect to the base of the robot; a homogeneous transformation that provides the position and orientation of the tool device relative to the device base; 30 a homogeneous transformation from the tool of the device at the base of the device at the beginning of trajectory recording.