Process for interacting with an object

Abstract

A process of interacting with an object (1a) is provided, comprising a robot (2) with an end effector (21), a drive (22) of the end effector (21), and a sensor (23) for acquiring at least one environmental parameter or interaction parameter; a computer (5) for controlling the robot (2); an instruction block (3) configured for instructing the robot (2) to move the end effector (21) according to instructions defined by the instruction block.

Description

Technical Field

[0001] This invention relates to the process of interacting with objects.

[0002] Specifically, the present invention relates to a process configured to preferably use robots to handle objects on automated production and assembly lines. Background Technology

[0003] It is well known that automated production and assembly lines are particularly flexible, thanks to the use of robots capable of performing different tasks and handling products of different shapes and uniformities.

[0004] These robots are now used to move objects between different workstations, such as moving objects between warehouses and workstations, and / or to perform tasks such as assembly and welding.

[0005] Through programming, they can faithfully perform repetitive tasks with great precision. These actions are determined by software, which specifies a series of coordinated movements in terms of direction, acceleration, speed, and distance.

[0006] These actions are defined by the operator, who must define instructions for each action and thus define the instructions issued to the robot to enable it to perform the required actions.

[0007] The existing technology described above has some significant drawbacks.

[0008] In particular, the known processing procedures require the use of specific programming languages ​​to guide the robot in completing each action for each new task. This leads to complex production process management and frequent errors due to the inability to accurately identify the correct process to be executed.

[0009] To address this issue, we designed a new instruction definition process. The new commercial robot is equipped with a graphical user interface (GUI), allowing operators to complete programming in a short time.

[0010] While this solution significantly speeds up programming, it doesn't offer sufficient improvement. Therefore, current object handling processes are particularly complex, difficult to use, relatively expensive, and / or require complex and costly robots.

[0011] Another drawback is the lack of flexibility in the known handling process, which almost always requires reprogramming because the robot cannot adapt to any changes in the actions to be performed and / or the objects to be handled. Summary of the Invention

[0012] In this context, the fundamental technical task of the present invention is to design a process for interacting with an object that can at least eliminate some of the aforementioned disadvantages.

[0013] Against the backdrop of the aforementioned technical tasks, an important objective of this invention is to obtain an object interaction process that is easy to program and therefore extremely flexible.

[0014] The technical tasks and prescribed objectives are achieved through the process of interacting with objects according to various embodiments of the present invention. Attached Figure Description

[0015] The features and advantages of the present invention will be explained below with reference to the accompanying drawings and through a detailed description of preferred embodiments, wherein:

[0016] Figure 1 A device configured according to the present invention for realizing the interaction process with an object is shown to scale;

[0017] Figure 2a Displayed to scale Figure 1 The assembly of equipment;

[0018] Figure 2b Displayed to scale Figure 2a Components for different purposes;

[0019] Figure 3 The environment for realizing the interaction process with objects according to the present invention is shown to scale; and

[0020] Figure 4 A schematic diagram illustrating the interaction process with an object according to the present invention is shown. Detailed Implementation

[0021] In this document, measurements, numerical values, shapes, and geometric references (such as perpendicularity and parallelism), when associated with “approximately” or other similar terms (such as “around” or “roughly”), should be considered, in addition to measurement errors or inaccuracies caused by production and / or manufacturing errors, primarily as slight deviations from the numerical value, measurement, shape, or geometric reference with which they are associated. For example, if these terms are associated with a numerical value, it is preferable to indicate a deviation of no more than 10% of that value.

[0022] Furthermore, when using terms such as "first," "second," "higher," "lower," "primary," and "secondary," it is not necessary to determine the order, hierarchy, or relative position of the relationship, but simply to clearly distinguish their different components.

[0023] Unless otherwise stated, the measurements and data reported herein should be considered as measurements and data taken in accordance with the International Civil Aviation Organization International Standard Atmosphere (ISO 2533:1975).

[0024] Unless otherwise stated, in the following discussion, "processing," "calculating," "determining," "counting," or similar terms refer to the operation and / or process of a computer or similar electronic computing device that can operate and / or convert data expressed as physical quantities, such as electronic quantities in registers in computer systems and / or memory, as well as other data expressed as physical quantities in computer systems, registers, or other storage, transmission, or information display devices.

[0025] Referring to the accompanying drawings, the process of interacting with an object according to the present invention is generally indicated by reference numeral 1.

[0026] Process 1 is configured to identify and interact with an object 1a in an environment 10 (e.g., a house or apartment). Preferably, it is configured to identify and interact with an object 1a in a work / industrial environment 10 (e.g., a warehouse or production line) to move / carry the object 1a.

[0027] Environment 10 can define a walkable area 10a.

[0028] Environment 10 may include at least one room 11. Specifically, it may include multiple rooms 11 and at least one passageway 12 between each room.

[0029] The interaction process 1 may include at least one robot 2.

[0030] Robot 2 can be configured to interact with at least one object 1a. Robot 2 may include at least one end effector 21 for grasping at least one object 1a.

[0031] The end effector 2 can interact with object 1a, particularly with multiple objects 1a of appropriately different weights and / or shapes. For example, it is configured to interact with only one object 1a at a time and precisely manipulate / perform grasping on it.

[0032] The end effector 2 may include at least two mechanical fingers that can move relative to each other in order to perform the grasping operation described above. It preferably consists of two or more fingers, especially recognizable in robotic hands (such as the robotic hands described in US2019269528 and US2018311827).

[0033] For each end effector 21, the robot 2 may include a driver 22 for the end effector 21.

[0034] The driver 22 can be configured to move the end effector 21 relative to the object 1a and / or the environment 10.

[0035] The actuator 22 may include a robotic arm. It may include one or more rigid bodies 221 (identifiable by a selective telescopic profile) and one or more joints 222 adapted to move the rigid bodies 221 in a suitable independent manner.

[0036] The connector 222 can be configured to rotate the rigid bodies 221 relative to each other by changing the unfolding angle between the two rigid bodies 221.

[0037] The joint 222 preferably moves the rigid body 221 according to inverse kinematics or direct kinematics. Therefore, even if not explicitly stated, each action of the robot 2 and the instructions defining these actions are determined according to inverse kinematics or direct kinematics.

[0038] The term "inverse kinematics" defines the trajectory in the operational space of the path of the end effector 21. Therefore, the velocities and accelerations of the individual joints 222 are defined, thereby forming the end effector path 21.

[0039] The term "direct kinematics" refers to the calculation of spatial trajectories, which determines the position, velocity, and acceleration of each joint 222, rather than the path of the end effector 21. Therefore, the path of the end effector 21 is the result of the position, velocity, and acceleration of each joint 222.

[0040] Each connector 222 can be electrified, specifically including a servo motor.

[0041] The actuator 22 may include a displacement device 223 configured to move the robot 2 along a walkable plane 10a.

[0042] The displacement device 223 may be electrically powered. For example, it may include at least one track or wheel.

[0043] Robot 2 may include sensor 23 for acquiring at least one suitable environmental parameter and / or for interaction between robot 2 and object 1a.

[0044] Sensor 23 may include one or more sensors, each configured to acquire a parameter selected from environmental parameters and interaction parameters. It is preferably configured to acquire at least one environmental parameter and at least one interaction parameter.

[0045] The term "environmental parameters" refers to parameters external to robot 2 and are therefore unrelated to the operation of robot 2. Environmental parameters can be physical conditions specific to environment 10 (such as temperature, humidity, and brightness) and / or characteristics of object 1a (such as shape and / or weight).

[0046] The term "interaction parameters" refers to parameters related to the operation of robot 2 when interacting with / manipulating object 1a. For example, the gripping position and / or force of end effector 21 or the contact temperature of end effector 21 with object 1a can be identified.

[0047] When acquiring at least one environmental parameter, the sensing device 23 may include one or more sensors, which may be selected from thermometers, photodetectors, hygrometers, or devices for photographing object 1a (such as cameras).

[0048] In the case of interactive parameters, sensor 23 may include one or more sensors, which may be selected from thermometers, piezoelectric sensors, and encoders for each connector 222.

[0049] Sensor 23 can be configured to acquire the robot's motion in order to define the necessary instructions to determine the robot 2's motion.

[0050] Robot 2 may include card 24 for controlling robot 2, particularly controlling at least end effector 21 and drive 22.

[0051] Card 24 can connect to sensor 23 for data transfer.

[0052] The interaction process 1 may include instruction block 3 for defining and sending instructions to robot 2 to move end effector 21 (preferably driver 22) to interact with one or more objects 1a.

[0053] Instruction block 3 can connect to robot 2 via data, specifically to computer 24.

[0054] Block 3 can be configured to define and send necessary instructions to robot 2 to determine the position of object 1a and perform interactions (handling, grasping, and / or moving) on ​​object 1a in environment 10. For example, block 3 can be configured to instruct robot 2 to locate object 1a present in environment 10, grasp object 1a, perform one or more operations on said object 1a, and then store it at a location in environment 10.

[0055] Instruction block 3 may include a computer or other device configured to allow an operator to input the instructions.

[0056] Alternatively, instruction block 3 can be at least partially wearable ( Figure 1 This allows the operator to simulate the operation and then send the instructions to the robot 2 based on the operator's actions. It may include a visualization device 31 configured to display at least some parameters collected by the sensor 23 to the operator; and a data acquisition device 32 configured to acquire the actions performed by the operator and send the necessary instructions to the robot 2 so that the robot repeats the aforementioned actions.

[0057] The data acquisition device 32 may include a sensing glove configured to detect the operator’s hand movements and send instructions to the robot 2 to allow the end effector 21 to perform the hand movements.

[0058] Additionally, device 32 may include a sensing device configured to detect the operator's actions and then send instructions to robot 2 to perform the actions.

[0059] Instruction block 3 may include at least one camera 33, and viewer 31 may include a screen (e.g., virtual glasses) for viewing images captured by the camera.

[0060] Camera 33 can be integrated into robot 2. Figure 1 ).

[0061] Alternatively, camera 33 can also be integrated into room 10. Instruction block 3 preferably includes at least one camera 33 in each room 11, the camera 33 being configured to capture objects 1a and / or robots in the room.

[0062] The interaction process 1 may include at least one marker 4, which can be detected by sensor 23 to identify the path of robot 2 by detecting the passage of robot 2 through specific points (e.g., between two rooms 11, between an object in a room, and / or between areas in the same room).

[0063] The interaction process 1 may include a marker 4 associated with each channel section 12 so that the sensor 23 can detect the passage of the channel section 12.

[0064] Alternatively or additionally, at least one tag 4 associated with each room 11 may be included so that the sensor 23 can detect whether the robot 2 has entered the room 11.

[0065] Alternatively, in addition to one or more markers 4, robot 2 can also detect its own path using sensor 23, which is configured to acquire and identify elements (such as furniture, doors, or object 1a) on the path. The robot then detects its own path based on its position relative to one or more elements detected by the sensor 23.

[0066] Process 1 may include robot 2 controlling computer 5. Alternatively, it may include multiple robots 2 and computer 5 controlling the robots 2.

[0067] Computer 5 can be configured to control end effector 21 (preferably driver 22) based on sensor 23 (i.e., one or more parameters acquired therefrom) and instructions described below.

[0068] It can connect to robot 2, circuit board 25 and / or sensor 23 for data communication.

[0069] Computer 5 can establish a data connection with instruction block 3.

[0070] Computer 5 can be configured to segment an action into a series of scenes based on a scene end instruction. In particular, it can segment an action into multiple scenes with each scene end instruction.

[0071] The scene end command can be issued automatically and / or manually.

[0072] In the case of an automatically issued scene end command, computer 5 may include a clock configured to measure the elapsed time during the execution of a real / virtual action (described below), and then issue a scene end command when the elapsed time is substantially equal to the scene duration.

[0073] In the case of a manually issued scene end command, computer 5 can issue the scene end command via block 3 after detecting an instruction performed by the operator (e.g., holding the position for at least a time threshold). Alternatively, block 3 may include a signaler (such as a button) that can be activated when the operator wishes to send a scene end command.

[0074] Alternatively, the robot 2 can detect the marker 4 or the sensor 23 can identify elements in the environment 10 to issue an instruction to automatically end the scene.

[0075] The computer may include the action database described below.

[0076] The interaction process 1 may include at least one learning stage 6, wherein the instruction block 3 defines a virtual action, in which the robot 2 moves the end effector 21, the computer 5 commands the robot 2 to perform an actual action according to the virtual action, and the sensor 23 acquires one or more parameters during the execution of the actual action.

[0077] Virtual actions include one or more instructions that determine the actions that robot 2 should learn during the learning phase based on the instructions received from instruction block 3.

[0078] Real-world actions include one or more movements performed by robot 2 according to virtual action instructions. The movements of real-world actions are executed in the order of the virtual action instructions. Therefore, real-world actions can be similar to, and optionally identical to, virtual actions.

[0079] The actual actions can be simulated (robot 2 does not move and / or manipulate one or more objects 1a) and / or real (robot 2 moves and / or manipulates one or more objects 1a). Real actions are preferred.

[0080] The interaction process 1 preferably includes multiple learning stages 6. Therefore, block 3 can define multiple virtual actions, which are preferably at least partially different from each other, thereby enabling the robot 2 to perform multiple real actions.

[0081] Each learning phase 6 may include a simulation sub-phase 61, where instruction block 3 defines the virtual actions of the robot 2's mobile end effector 21.

[0082] In simulation sub-stage 61, instruction block 3 defines virtual actions and sends instructions to robot 2 to execute the actual actions corresponding to the virtual actions.

[0083] In some cases, virtual actions may include instructions relating to the robot’s proper passage through at least one passageway section 12 between two or more rooms 11.

[0084] One or more scene end commands can be issued during simulation sub-stage 61.

[0085] Learning phase 6 may include a repetitive sub-phase 62, in which computer 5 commands robot 2 to perform actual actions based on the virtual actions.

[0086] In the repeating sub-stage 62, the computer 5 commands the robot 2 to perform one or more movements, which define the real movements based on the instructions of the virtual movements.

[0087] Preferably, in sub-stage 62, computer 5 commands robot 2 to substantially repeat the actions in the virtual actions. Therefore, the real actions are essentially the same as the virtual actions.

[0088] In the repeating sub-phase 62, robot 2 can move from one room 11 to another via at least one passage section 12 (if provided in the virtual action).

[0089] Learning phase 6 may include acquisition sub-phase 63, in which sensor 23 acquires one or more parameters during repeated sub-phase 62.

[0090] In sub-stage 63, sensor 23 may acquire other environmental parameters (such as the brightness, pressure and / or temperature of environment 10) and / or one or more interactive parameters, such as the gripping force of end effector 21 and / or the force applied thereto.

[0091] In the acquisition sub-stage 63, the sensor 23 preferably acquires at least environmental parameters of the object 1a (e.g., shape and / or color) to allow the robot 2 to detect the presence of the object 1a in the environment 10.

[0092] Additionally, sensor 23 may acquire at least one marker 4 to allow the computer to identify the executed path. Alternatively, in addition to one or more markers 4, sensor 23 may acquire one or more elements in the environment 10, and robot 2 may determine its own path based on its position relative to one or more elements detected by sensor 23.

[0093] In the acquisition sub-stage 63, sensor 23 can acquire the actions of robot 2, thereby defining the necessary instructions to determine the actions that robot 2 must subsequently perform, thereby repeating the aforementioned actions.

[0094] Finally, in the acquisition sub-stage 63, acquisition parameters, one or more scene end commands, and inscriptions related to the movement of robot 2 are sent to computer 5.

[0095] Sub-stages 61, 62, and 63 can be performed almost simultaneously.

[0096] The interaction process 1 may include an analysis phase 7, in which each actual action performed in the learning phase 6 is broken down into a series of scenarios.

[0097] In the analysis phase 7, the computer 5 can appropriately divide each real action into an initial scene and an ending scene according to at least one ending scene instruction, and preferably also have at least one scene between the initial scene and the ending scene.

[0098] Computer 5 can associate one or more instructions for robot 2 with each scene, these instructions defining the operations to be performed by the robot. The instructions are defined based on the motion of robot 2 detected in acquisition sub-stage 63.

[0099] In some cases, the initial scenario may be unrelated to the instructions.

[0100] Computer 5 can associate at least one value corresponding to the parameters (environment and / or interaction) obtained during the execution of the scene with each scene during the learning phase 6, specifically at the beginning of the scene, i.e. immediately after receiving the scene end instruction.

[0101] In this paper, the term "value" refers to the parameter recorded by sensor 23 in learning phase 6 and appropriately stored in the motion database below.

[0102] Computer 5 will preferably associate each parameter value acquired in the learning phase 6 with each scenario.

[0103] The interaction process 1 may include a reprocessing stage 8, in which the computer 5 creates an action database by associating actual actions with at least one value substantially the same as those associated with the initial scene, thereby defining a composite action consisting of a single initial scene and multiple final scenes.

[0104] Preferably, initial scenarios with the same values ​​are associated, and optionally, the actual operations of the same instructions are correlated.

[0105] The final number of scenes may be roughly equal to the number of actual actions combined together.

[0106] If there are one or more intermediate scenarios, two or more scenarios can only be merged if they have substantially the same instructions and optional identical values.

[0107] In summary, a compound action has only one initial scene, from which multiple final scenes branch out. If present, intermediate scenes can identify one or more branches connecting each final scene to the initial scene.

[0108] In the action database, each scene can be associated with one or more instructions from robot 2.

[0109] Each scene of the composite action can be associated with at least one value, namely, one or more parameters acquired by sensor 23 during scene execution in each learning phase 6 of the actual action, which is incorporated in the composite action. Specifically, when a scene is executed in different learning phases 6, for each parameter detected by the same sensor, if the parameters are nearly equal, the scene is associated with only one value; while if the parameters are different from each other, the scene is associated with multiple values, each corresponding to a parameter detected by the sensor in learning phase 6.

[0110] Each value of a scenario is associated with at least one subsequent scenario to determine the execution order of scenarios. In this document, the term "next scenario" refers to the scenario that is adjacent to or immediately following a given scenario.

[0111] If the same sensor detects different values ​​for a parameter, each value can only be associated with a subset of the following scenarios. Specifically, at least one scenario upstream of the bifurcation between alternative scenarios of a compound action can be associated with a value, and each value is associated with only a subset (specifically only one) of the subsequent alternative scenarios, so that robot 2 can determine which alternative scenario to execute based on the value.

[0112] The interaction process 1 may include an execution phase 9, in which the robot 2 executes an action from the action database in an appropriate automatic mode.

[0113] In detail, in execution phase 9, sensor 23 defines an initial factor by detecting the at least one parameter, computer 5 selects a compound action in the action database whose initial value is substantially equal to the initial factor, and commands robot 2 to perform an operation based on the scene following the initial scene of the identified compound action; computer 5 commands robot 2 to perform the next operation to be performed in the next scene associated with the selected compound action by comparing at least one value in the action database associated with one or more scenes before the next scene (and therefore still to be performed) with at least one factor obtained by detecting at least one parameter in one or more operations before the next scene.

[0114] In this paper, the term "factor" refers to the parameters recorded by sensor 23 in execution phase 9.

[0115] Execution phase 9 may include acquisition sub-phase 91, in which sensor 23 defines at least one initial factor by detecting the at least one parameter.

[0116] The initial factor is obtained when robot 2 starts up, that is, before it performs any actions.

[0117] Preferably, in the acquisition sub-stage 91, the sensor 23 acquires all parameters and then defines an initial factor for each parameter.

[0118] During the acquisition sub-stage 91, robot 2 can remain essentially stationary.

[0119] Execution phase 9 may include an identification sub-phase 92 for the compound operation to be performed.

[0120] In identification sub-stage 92, computer 5 selects a composite action from the action database that has an initial value substantially equal to the initial factor. Specifically, computer 5 selects a composite action from the action database that has an initial value substantially equal to all the initial factors defined in sub-stage 91.

[0121] After determining the compound action, computer 3 will end development sub-phase 93 and command robot 2 to execute the scene after the initial scene;

[0122] The execution phase 9 may include at least one development sub-phase 93 in which the robot 2 performs operations based on the identified scenario, and in particular based on the instructions associated therewith.

[0123] Therefore, the execution phase 9 may include at least one selection sub-phase 94, in which the computer 5 identifies the next scene to be executed in the scene that constitutes the action, and then commands the robot 2 to perform the operation according to the next scene.

[0124] In this sub-phase 94, computer 5 identifies the next scenario that will be appropriately executed in the following development sub-phase 93.

[0125] This selection can be achieved by comparing at least one value from at least one scenario preceding the next scenario with one or more factors defined in the operations performed so far. The next scenario to be executed may be a scenario connected to one or more values ​​from at least one preceding scenario, which are substantially equal to one or more factors corresponding to the parameters detected in the preceding operations.

[0126] You can start with the initial action and scene, and then perform this value-factor comparison on all actions and scenes.

[0127] After determining the next scenario, computer 5 commands robot 2 to perform operations according to the next scenario, and then executes the new development sub-stage 93 and the new selection sub-stage 94.

[0128] When the final scene is selected as the next scene in the selection sub-stage 94, and then the final operation is performed based on the final scene in the next development sub-stage 93, the execution stage 9 ends.

[0129] To illustrate the process, we will use an example to demonstrate the application of the interaction process 1.

[0130] Initially, the process was envisioned to perform two learning phases 6: a first phase 6 consisting of a first sub-phase of simulation 61, repetition 62, and acquisition 63, and a second learning phase 6 consisting of a second sub-phase of simulation 61, repetition 62, and acquisition 63.

[0131] In the first simulation sub-phase 61, the operator simulates the first virtual action via instruction block 3, including locating the glass, picking up an empty glass from the first room 12, moving to the second room 12, filling the glass with a carafe, and placing the glass on a tray. At the end of each part of the virtual action, the operator issues a scene end instruction; in this example, five scenes are defined.

[0132] Simultaneously with the first sub-stage 61, there is also a first repeating sub-stage 62, in which robot 2, guided by computer 5, repeats the above actions according to the signals received from block 3; and a first acquisition sub-stage 63, in which sensor 23 preferably sequentially acquires and identifies parameters of the glass; glass picking instructions and the weight of the empty glass; room replacement instructions 12 (identifying one or more elements and / or one or more markers 4); pot identification parameters and glass filling instructions; and finally, tray identification parameters (object placement position 1a) and instructions to place the glass on the tray.

[0133] After completing the first learning phase 6, the process stipulates that you will proceed to the second learning phase 6.

[0134] In the second simulation sub-phase 61, the operator simulates a second virtual action via instruction block 3, including: finding a glass, picking up a full glass from the first room 12, and arriving at the second room 12 and placing the glass on a tray. At the end of each part of the second virtual action, the operator issues a scene end instruction; four scenes are defined in the second example.

[0135] Simultaneously with the second sub-stage 61 is the second repetition sub-stage 62, in which robot 2, guided by computer 5, repeats the above actions based on signals from block 3; and the second acquisition sub-stage 63, in which sensor 23 preferably sequentially acquires glass identification parameters, glass picking instructions and full glass weight, room replacement instructions 12, and tray identification parameters and instructions to place the glass on the tray.

[0136] After completing all stages 6, the interaction process 1 will enter the analysis stage 7. In this stage, the two actions will be divided into multiple scenes according to the scene end instruction, and relevant instructions and parameters will be obtained in each scene.

[0137] The first real action is subdivided into the initial scene of positioning the glass, the first intermediate scene of picking up the empty glass from the first room 12, the second intermediate scene of arriving at the second room 12, the third intermediate scene of filling the glass with the Calaf bottle, and the final scene of placing the glass on the tray.

[0138] The second real action is subdivided into the initial scene of positioning the glass, the first intermediate scene of picking up the full glass from the first room 12, the second intermediate scene of arriving at the second room 12, and the final scene of placing the glass on the tray.

[0139] At this point, the scene needs to be reprocessed in stage 8, and then an action database needs to be created.

[0140] In this example, since the initial scene is the same, a compound action containing two final scenes is defined, and the branch is specifically described at the end of the second intermediate scene.

[0141] Specifically, the compound action provides an initial scenario associated with the "glass recognition" value; a first intermediate scenario associated with the pick-up instruction and two values ​​(full glass weight and empty glass weight); and a second intermediate scenario associated with the room 12 change instruction (detecting one or more markers 4 and / or one or more elements in the environment 10). Therefore, the compound action provides branches (i.e., the possibility of performing different operations). In the first branch, the compound action includes a third scenario related to filling the glass and a final scenario of placing the glass on the tray; while in the second branch, the compound action only includes the final scenario of placing the glass on the tray.

[0142] The first branch is associated with the weight of the empty cup in the first intermediate scene, and the second branch is associated with the weight of the full cup in the first intermediate scene.

[0143] At this point, we can proceed to execution phase 9.

[0144] In the acquisition sub-stage 91, sensor 23 detects / identifies the glass to determine the corresponding initial factor.

[0145] In identification sub-stage 92, computer 5 searches the action database for a composite action with an initial scene whose initial value is substantially equal to the initial factor, and commands the execution of the scene following the initial scene, i.e., the first intermediate scene. It is important to note that sub-stages 91 and 92 only end when an initial factor corresponding to the initial value existing in the action database is detected.

[0146] Then comes development sub-stage 93, where, based on the first intermediate scenario, robot 2 performs the operation of picking up a glass (e.g., a full glass), and sensor 23 acquires factors related to the weight of the glass.

[0147] At this point, computer 5 will search for scenarios related to the obtained factors in the scenarios following the first intermediate scenario (i.e., in one or more second intermediate scenarios), which is selection sub-stage 94. In this case, since only second intermediate scenarios exist, computer 5 issues an execution command.

[0148] Then, execution phase 9 provides a new execution sub-phase 93, in which robot 2 performs operations according to the second intermediate scenario. Then, moving from the first room 12 to the second room 12, sensor 23 collects collected factors corresponding to marker 4 and / or detected elements.

[0149] Since the final scene is not yet complete, a new selection sub-stage 94 will occur, in which computer 5 will search for scenes associated with the collected factors in the scenes following the second intermediate scene.

[0150] Since there are two possible consecutive scenarios, computer 3 searches among the factors collected in previous operations to see which scenario's (i.e., the second intermediate scenario, the first intermediate scenario, and the initial scenario) value corresponds to the factor collected in the corresponding sub-stage 93. Specifically, by identifying the detection of the "full glass" factor in the first sub-stage, it detects that the corresponding value is associated with the final scenario of the second case, identifies the final scenario of the second case as the next scenario, and commands the execution of that scenario.

[0151] Execution phase 9 provides a new sub-phase 93 for performing operations based on the final scenario. Specifically, robot 2 locates the tray in room 12, identifies the tray, and places the glass back on the tray.

[0152] When this development sub-phase 93 ends, phase 9 also ends, because the final scenario of the identified composite action has been executed.

[0153] The interaction process 1 of the present invention has significant advantages.

[0154] In fact, process 1 can issue commands to robot 2 in a practical and fast way, and most importantly, it can quickly update the database of specific actions.

[0155] Another advantage related to the creation and use of the operational database is that the interactive process 1 can manage a variety of different operations without using complex and expensive structures, and without errors in identifying the correct operation to be performed.

[0156] This advantage can also translate into the possibility of controlling multiple robots 2 simultaneously in a precise, simple, and fast manner.

[0157] Another advantage is that Robot 2's instructions are simple.

[0158] Therefore, another advantage lies in the high flexibility of the interaction process 1. In fact, it does not require constant reprogramming because the special use of sensor 23 enables robot 2 to adapt to changes in the actions to be performed and / or the object 1a to be interacted with. For example, robot 2 can recognize object 1a regardless of its position in environment 10 and / or the location, thus enabling it to quickly adapt to any working conditions and move and locate object 1a at the fastest speed.

[0159] This invention may be modified within the scope of the inventive concept defined in the claims.

[0160] For example, at least in interaction process 1, card 24 and computer 5 can overlap.

[0161] Furthermore, when selecting the next scenario in the selection sub-stage 63, the computer may select the scenario that is closest to the value of a factor obtained in at least one previous operation as the next scenario.

[0162] In some cases, the instructions may only determine the final position of object 1a and / or robot 2. Therefore, robot 2 will automatically determine the actions to be performed to achieve the final position based on the parameters collected by sensor 23. For example, the instructions may determine the position of object 1a on the field, and then robot 2 will decide on its own which actions to perform to place object 1a on the field.

[0163] In some cases, sensor 23 may include a camera, and display screen 31 may display the camera's image.

[0164] In such an implementation scheme, all details can be replaced with equivalent components, and materials, shapes, and sizes can be arbitrarily chosen.

Claims

1. An interaction process with an object (1a), comprising: - At least one robot (2), which includes An end effector (21) that interacts with at least one of the objects (1a). The driver (22) of the end effector (21); A sensor (23) for acquiring at least one parameter; the at least one parameter includes at least one of environmental parameters and parameters of the robot's interaction with the object (1a); - A computer (5) for controlling the robot, which is connected to the sensor (23) for data communication; - Instruction block (3), which instructs the robot (2) to move the end effector (21); - Multiple learning stages (6); in each learning stage The instruction block (3) defines a virtual action in which the robot (2) performs the action of the end effector (21); The computer (5) executes real actions on the robot (2) according to the virtual action command, and The sensor (23) acquires at least one parameter when performing the actual action; -Analysis phase (7), wherein the computer (5) is described. Each of the aforementioned real actions is further subdivided into an initial scene and a final scene, and Associate at least one value corresponding to at least one parameter collected in the scene with each of the scenes; - Reprocessing stage (8), in which the computer (5) creates an action database that associates real actions with the same initial value in order to define a composite action consisting of a single initial scene and multiple final scenes; the action database associates at least one value with each scene; - Execution phase (9), where The sensor (23) defines at least one initial factor by detecting the at least one parameter; The computer (5) selects a composite action with an initial value equal to the initial factor from the action database and commands the robot (2) to perform an operation equal to the scene following the initial scene of the composite action; as well as The computer (5) commands the robot (2) to perform the next operation, which is selected from the scenarios related to the composite action in the action database; The selection is made by associating at least one value from the action database with at least one scene preceding the next scene. Compare with at least one factor obtained by detecting at least one parameter in an operation prior to the subsequent operation; The characteristic is that the execution phase (9) includes - Acquisition sub-stage (91), wherein the sensor (23) defines the initial factor by detecting the at least one parameter. - Identification sub-stage (92), wherein the computer (5) selects an action with an initial value equal to the initial factor from the action database and commands the robot (2) to execute at least one scene after the initial scene; - At least one development sub-stage (93), wherein the robot (2) performs an operation based on the initial scenario of the composite action selected in the recognition sub-stage (92), and defines a factor by acquiring at least one parameter. - At least one selection sub-stage (94), wherein the computer (5) identifies the next scene to be performed in the initial scene of the compound action and commands the robot (2) to perform an operation according to the next scene; and wherein the computer (5) identifies the next scene by comparing at least one value associated with at least one scene preceding the next scene and at least one factor collected in a previous operation; The sensor (23) is configured to acquire the movements of the robot (2) by means of instructions required to define the movements of the robot (2); and In the reprocessing stage (8), the computer (5) associates the instruction with at least one scenario in the action database.

2. The interaction process according to claim 1, wherein the initial scene does not contain the instruction.

3. The interaction process according to claim 1, wherein when one of the final scenes is selected as the next scene, the execution phase (9) terminates and then the robot (2) is commanded to perform a final operation according to the final scene.

4. The interaction process according to claim 1, wherein at least one scene end instruction is defined in each of the learning stages; and wherein in the analysis stage (7), the computer (5) subdivides each of the real actions into the initial scene and the final scene according to the at least one scene end instruction.

5. The interaction process according to claim 1, wherein at least one scene end command is automatically issued at each learning stage.

6. The interaction process according to claim 5, wherein the computer (5) includes a clock configured to measure the elapsed time; and wherein in each learning phase (6), the computer (5) defines at least one scene end instruction when the elapsed time is equal to the scene duration.

7. The interaction process according to claim 5, wherein in each learning phase, at least one scene end command is issued manually.

8. The interaction process according to any one of claims 5-7, wherein the instruction block (3) includes a signaler configured to send the scene end instruction.

Images