FR3142318A1 - Robotic network architecture with a set of labels - Google Patents

Abstract

The present invention relates to a robotic network architecture comprising a set of labels, at least one robot comprising at least one predefined gripper, said labels (30) being configured to be respectively connected to an element belonging to the robotic network and containing information of identification of said element and/or location information of said element and/or characteristic information of the element and/or information on the ability of the element to be grasped and/or manipulated and/or moved by the robot means of the predefined gripper, the elements being objects and/or people and/or places, the robot being configured to interact with any element comprising at least one label (30) and said robot comprising means for interpreting said information identification and/or location and/or characteristic and/or the ability of the element to be grasped and/or manipulated and/or moved by the robot (10), as well as means for determining a specific behavior -with respect to said element, in particular with regard to seizure and/or manipulation and/or movement. Abstract Figure: Figure 1

Description

Robotic network architecture with a set of labels

The present invention relates to the field of robots, in particular humanoid robots. More specifically, the invention relates to a robotic network architecture comprising a set of labels, at least one robot comprising at least one predefined gripper, said labels being configured to be connected to an element and the robot being configured to interact with any type element, in particular any type of object, comprising at least one label.

STATE OF THE ART

Robots with a single spherical wheel on which the entire robot rests are known. These types of robots are commonly called “Ballbots”. Said robots are mobile by means of the single spherical wheel, in all directions on a flat surface, in a stable manner.

In particular, robots with a single spherical wheel are known, each comprising a trunk, two arms and an upper part. The trunk is connected to the spherical wheel and it is fixed relative to it, each arm is connected to the trunk by means of a pivot connection or by means of a ball joint so as to allow the mobility of each of the arms relative to the trunk and the upper part is connected to an upper portion of the trunk and it is fixed relative to the latter. The upper part of the robot includes at least one sensor configured so that the robot finds its way in its environment. Such a robot is configured to move in all directions and to perform tasks using its arms, such as grabbing objects.

To grasp objects, said robots include a gripper. The gripper is configured to allow the robot to grasp objects, regardless of their shape.

Gripping systems specific to an object are known, that is to say that there are robots provided with a gripper configured to grasp a single type of object, for example, the robot is provided with a suction cup configured to allow the robot to grasp a glass plate.

However, such gripping systems make the robot single-task and specific to a single type of object.

To overcome the aforementioned problem, there are robots with gripping systems configured to grasp objects, whatever their shape. In particular, such gripping systems consist of robotic hands configured to reproduce the movements of a human hand. Such systems have a large number of parts and great mechanical complexity.

In addition, it is known to add a visual marker on the objects to be grasped to allow the robot to identify an object, by image recognition using image processing software and/or neural network algorithms, For example. This type of visual marker is suitable to be detected and processed by the robot. However, such detection and the associated processing are not possible if the visual marker is not visible, if the brightness is too low or if the robot is too far away. In addition, these known techniques are not always reliable, because the algorithms and software implemented may fail to recognize an object for example.

The present invention therefore aims to resolve the aforementioned problems by proposing a robotic network architecture comprising a set of labels, at least one robot comprising at least one predefined gripper, said labels being configured to be connected to an element and the robot being configured to interact with any type of element comprising at least one label, so as to allow the robot to recognize any element equipped with a label, including in particular in the event of low light, to detect and identify an element which is not there, at least initially, in the field of vision of the robot etc.

PRESENTATION OF THE INVENTION

More precisely, the subject of the invention is the establishment of a robotic network architecture comprising a set of labels, at least one robot comprising at least one predefined gripper, said labels being configured to be respectively connected to an element belonging to the robotic network and containing identification information of said element and/or location information of said element and/or characteristic information of the element and/or information on the ability of the element to be grasped and/or manipulated and/or moved by the robot by means of the predefined gripper, the elements being objects and/or people and/or places, the robot being configured to interact with any element comprising at least one label and said robot comprising means of interpretation of said identification and/or location and/or characteristic information and/or the ability of the element to be grasped and/or manipulated and/or moved by the robot, as well as means for determining a behavior with respect to said element, in particular in terms of seizure and/or manipulation and/or movement.

Advantageously, the robotic network architecture further comprises a set of universal gripping means devices configured to be graspable by means of the predefined gripper, the universal gripping means devices being intended to be connected respectively to a set of objects of the set of elements, the universal gripping means devices comprising at least one of said labels and said universal gripping means devices being configured to be grasped and/or manipulated and/or moved by means of the predefined gripper.

According to one embodiment, at least part of said labels of said set of labels each comprise a capsule, a transmitting antenna and a battery, the transmitting antenna and the battery being housed in said capsule, and in that said robot comprises a receiving antenna configured to communicate with said transmitting antenna.

In particular, said universal gripping means device comprises a body and at least one fixing means, the body being configured to be graspable by the predefined gripper and the fixing means being configured to connect the body to said object, said body extending mainly in a longitudinal direction between a first end and a second end, and the capsule being housed in a housing provided at one of said ends of the body.

Advantageously, at least part of said labels further comprise at least one button arranged on an external face of the capsule and configured to activate/deactivate the label and/or to call the robot and/or to send a predefined message to the robot.

Advantageously, at least part of said labels further comprise at least one inertial unit configured to detect a movement or an orientation of the elements which are equipped with it.

Advantageously, at least part of said labels also comprise a Bluetooth chip coupled to the transmitting antenna.

According to one embodiment, the bluetooth chip meets the 5.1 standard or higher and includes a direction finding function making it possible to determine an angle of arrival in relation to an object also equipped with a bluetooth chip meeting the 5.1 standard or higher and including a compatible direction finding function.

PRESENTATION OF FIGURES

The invention will be better understood on reading the description which follows, given solely by way of example, and referring to the appended drawings given by way of non-limiting examples, in which identical references are given to similar objects. and on which:

There represents a label, in the form of a tag presenting a capsule, in isometric view;

There shows a tag presenting a capsule, in profile view;

There is a schematic representation seen from the front, of a tag comprising a capsule and external electrical connectors;

There represents a universal gripping means device comprising a tag;

There is a schematic representation of a robot capable of interacting in the robotic network architecture.

It should be noted that the figures set out the invention in detail to enable the invention to be implemented; although not limiting, said figures serve in particular to better define the invention where applicable.

DETAILED DESCRIPTION OF THE INVENTION

The invention relates to a robotic network architecture comprising a set of labels and at least one robot comprising at least one predefined gripper. The labels are configured to be linked to an element and contain information identifying the element, as well as information allowing the location of the element and/or other information allowing the robot to determine the behavior to adopt towards the element. Thanks to the information contained in the label, the robot can in particular determine whether it has the possibility, in particular the capacity and/or the right, to grasp and/or manipulate and/or move, in particular by means of the predefined gripper, said element. The robot is configured to interact with any type of element comprising at least one label and said robot includes means for interpreting said identification information, as well as other information possibly contained in the label.

The elements are in particular objects, but they can also be people or places in particular.

With reference to Figures 1 and 2, a label 30, also referred to as a “tag”, is primarily a means of identifying elements of the environment in the robotic network. To this end, the tag 30 includes identification information for the element with which it is associated. This identification information is for example a unique identifier or a name. The tag 30 allows the robot(s) elements of the robotic network to detect the presence of the equipped elements, and to name them. The tag 30 can also allow the robot to determine the position of the equipped elements. The tag can be a label attached to the element, for example glued, and including a color, a code, a particular pattern that the robot can decode by image analysis.

Also, the tag 30 may include a capsule containing means of communication with the robot, so as to trigger an action on its part, to indicate a priority of action, to transmit data from sensors carried by the element, etc. . For these purposes, the tag may include a memory and a bluetooth antenna type transmitter antenna.

According to one embodiment, a tag 30 may in particular contain information linked to the element with which it is associated, such as its descriptive name, in addition to an identifier, in particular a unique identifier. In particular, such information may include the status of the person equipped with a tag, giving him, for example, administrator status authorizing him to give more instructions that the robot must execute. If the person equipped with a tag is not an administrator, based on the information in the tag they carry, then certain instructions may not be executed by the robot.

The tag 30 may include information allowing the robot to locate the element carrying it or a surrounding element, either by direct location information, or via the AOA bluetooth functionality described below.

Optionally, the tag 30 may include information corresponding to characteristics of the element which carries it, such as its size, its weight, or the behavior to be adopted by the robot in terms of seizure and/or manipulation and/or movement etc.

This information can be transmitted to the robot via means of communication, in particular a Bluetooth antenna housed in a capsule forming part of the tag 30, in order to allow the identification of the element by the robot and to facilitate its interaction with it.

A tag 30 can also be connected to an external device 31, 32, in particular via I2C pins provided on the back of the tag, and thus transmit the information contained therein to the robot. For example, the tag 30 may include an external electrical connector 32, configured to allow recharging of the battery of the tag, or even of the object equipped with the tag, or an external electrical connector 101 with four pins, configured to allow connection of a device external to the tag, such as a sensor, a communication device, etc.

According to one embodiment shown in Figures 1 to 4, a tag 30 comprises a capsule to house any means of communication and memory, as well as a button allowing a signal to be transmitted to the robot. For example, a person can activate this button present on a tag attached to an element of the robotic network, when the robot is nearby, that is to say within a radius allowing the robot to detect the tag in question.

The button can make it possible, for example, to activate or deactivate the tag 30, to send a signal to call a robot, to give the element which is equipped with it a particular character, such as being an object of interest for the robot, or a temporarily prohibited area, etc.

The button can also be used to call the robot or send it a predefined message.

As already indicated, a tag can be associated with different categories of elements, such as a handle for gripping an object, an area, which can be identified as a drop zone for example, in a building, on a wall or a door, etc., thus allowing the robot to identify these elements on a semantic level and to determine the behavior it can adopt towards them.

A tag 30 can also be worn by a person, thus allowing the robot to identify them and interact with them.

A tag 30 can in particular be used to define an area of interest in the environment in which the robotic network architecture is deployed, giving rise to a specific instruction for the robot, by means of information contained in the tag 30 and interpretable by the robot, such as: prohibited zone, authorized zone, reduced speed zone, etc.

In the robotic network architecture, the robot can thus be configured so as to understand a database making it possible to associate a tag 30, in other words information contained in the tag, with a type of element with which it is associated. Depending on the type of element, the robot will be configured to adopt appropriate behavior.

For example, an element of type "handle" may be draggable, but an element of type "area" or an element of type "person" are not. On the other hand, the robot could potentially receive instructions from a “person” type element, for example provided that it is an authorized person, depending on the instructions received, etc.

In particular, the tag 30 includes a battery, also housed in the capsule, if applicable. The battery is sized to be able to provide electrical power to the tag 30, preferably for at least several months.

According to the invention, the possibility of “teaching” the robot to recognize and interpret a new tag 30 associated with an element of the robotic network architecture is provided. For example, to define a new association of a tag 30 with an element, the user can expose the tag 30 near the robot and associate it with a semantic description, such as: “this tag is a bottle of water”. The identification of tag 30 is stored in a robot database, and associated with the name “water bottle”. If the tag 30 is associated with a handle, the robot will know that it can move said “water bottle” because the tag 30 includes the information interpretable by the robot, according to which the element associated with the tag 30 is a “handle” type element and that this type of element is movable, said handle being in practice fixed or forming an integral part of the water bottle in question.

1. « prends la bouteille d’eau, et dépose-la sur la table dans la zone de dépôt » ;
2. « suis-moi » ;
3. « va dans la zone A » ;
4. etc.

Thus, in operation of the robotic network, once elements have been associated with a tag 30 (handle, drop zone, etc.), instructions can be given to the robot with reference to the tags, such as for example:

1. “take the water bottle, and place it on the table in the drop-off area”;
2. " follow me ";
3. “go to zone A”;
4. etc.

According to one embodiment, it is planned that the batteries of the tags 30 are rechargeable, in particular via a charging dock capable of recharging several tags simultaneously.

According to the embodiment illustrated in Figures 1 to 4, the tags 30 are composed of a Bluetooth transmitter having the AOA functionality (for Angle Of Arrival meaning angle of arrival), which allows the robot, itself equipped with an AOA bluetooth receiving antenna, to determine the direction in which said tags 30 are located. Thus, the AOA bluetooth chip of the tag 30 is configured to communicate with the robot and to indicate to it the angular position of the tag 30 and, for example, its unique identifier.

In a robotic network architecture comprising several fixed or mobile elements equipped with such tags 30, the robot thus carries out several measurements at different positions of the robot. The robot can thus include means for determining, by radio direction finding, the position of each of the tags 30 by triangulation of the bluetooth signal emanating from each of said tags 30, thanks to the AOA functionality and the adapted determination means.

The tag 30 can also include, housed in the capsule, an inertial unit making it possible to detect the movement of the element equipped with it. Such an inertial unit can also be configured to allow good orientation of the element associated with tag 30, in particular when the element is grasped, manipulated, moved by the robot, or to help identify the position of the corresponding element.

The tag 30 can also be associated with one or more sensory sensors.

The tag 30 constitutes in particular a unique means of semantic recognition of the element which is equipped with it, from the point of view of the robot. In other words, the tag constitutes in particular unique identification information preferably associated with other information relating to the element which is equipped with it, including structural and/or functional characteristics, in particular. Thanks to the tags 30, the elements which are equipped with them, and which are external to the robot, form part of the robotic network within which the robot will be able to determine the attitude that it can adopt towards these elements.

For example, items are classified into different types, including objects, people and places. Subtypes can be defined, for example to form classes of objects, places or people. This makes it possible to simplify the categories of possible interactions between the robot and these elements, depending on whether the element belongs to a type of elements (place, person or object) and, for a type of element, to a class.

For example, the robot will not consider an area, a person, or an object belonging to the class of prehensile objects as equivalent. A tag 30 on an area makes it possible to identify a place to which the robot can or cannot move, a room into which the robot can enter or not. Or, a tag 30 on an object can be used to indicate to the robot that the object can be grasped and movable or not.

Thus, more generally, the tag 30 makes it possible to open a generative space for the robot in the sense that the existence of the tags 30 in the robotic network architecture creates an environment making it possible to automate predefined tasks, such as making a roll. between different objects at given times, bring objects back to the previously defined point when the objects were moved etc. It is because these objects are elements equipped with tags 30, that the people giving instructions are equipped with tags authorizing them, and that the places between which objects are moved are elements equipped with tags, that this is possible .

As indicated previously, in the robotic network architecture according to the invention, at least one robot is provided adapted to interact with elements equipped with tags.

In particular, the robot can be a robot 10 comprising a single spherical wheel 2 intended to be in contact with the ground and a platform 3 mounted on the spherical wheel 2 by means of stabilizing means, as shown in the figure. . Said robot 10 rests on the spherical wheel 2 via the platform 3. This type of robot 10 is commonly called a “Ballbot”.

The robot 10 is mobile on the ground by means of the single spherical wheel 2, in all directions, in a stable manner.

The robot 10 comprises a trunk 5, at least one arm 4 and an upper part 7.

In particular, the trunk 5 is connected on the one hand to the platform 3 and on the other hand to the upper part 7 and the at least one arm 4 is connected by means of a pivot type or ball joint type connection to the trunk 5.

The robot 10 may include a plurality of arms 4, preferably two arms, like a humanoid robot. As visible on the , the robot 10 thus comprises two arms 4, on each side of the trunk 5, each arm 4 having a portion forming a forearm and a portion forming a rear arm, with a joint between the two. At the end of each arm 4, the robot includes a gripper 6, notably resembling a hand. The gripper 6 is preferably predefined and can have any shape and function: hand, gripper, suction cup, etc.

The robot 10 preferably has an upper part 7 forming a head equipped with an optical sensor, in particular a camera.

According to one embodiment of the robotic network architecture, part of the elements are associated with universal gripping means devices dedicated to the robot's grippers and comprising a tag. Each universal gripping device is intended in particular to be connected to “any type” of object and it includes a tag 30 containing identification information of the object and possibly other information, as described previously.

The gripper 6 of the robot 10 is thus adapted to interact with said elements of suitable type, in particular elements of the "object" type, via the universal gripping means device 1, integrating a tag 30 or to which is attached a tag. In particular, the universal gripping means device 1 is specially configured to be easily graspable by means of the predefined gripper 6 of the robot 10.

For example, such a universal gripping means device is as shown in the . The universal gripping means device 1 is configured to be grippable by means of a predefined robotic gripper 6. The universal gripping means device 1 is in particular intended to be connected to a set of objects and comprises a tag 30 which is housed in a housing at one end of the universal gripping means device 1.

A universal gripping means device 1 comprises for example a body and at least one fixing means, the body being configured to be graspable by the predefined gripper 6 and the fixing means being configured to connect the body to said object, said body s extending mainly in a longitudinal direction between a first end and a second end, and the capsule being housed in a housing provided at one of said ends of the body.

According to one embodiment, the robotic network architecture includes a regularly updated database. The database contains, for example, the location of the tags 30. It is updated during the movements of the robot 10. The robot 10 is in particular configured to detect, using an on-board antenna, the presence, the direction and to estimate the distance of the tags. 30 assets nearby. The robot 10 can then be configured to update the database during its movements within the robotic network, which allows it to move towards a given tag 30 if it receives the instruction. Once nearby, the location will allow the robot 10 to direct itself and approach said tag 30 at an interaction distance.

The invention is not limited to the embodiments previously described but extends to any equivalent embodiment.

Claims (8)

Robotic network architecture comprising a set of labels, at least one robot (10) comprising at least one predefined gripper (6), said labels (30) being configured to be respectively connected to an element belonging to the robotic network and containing information identification of said element and/or location information of said element and/or characteristic information of the element and/or information on the ability of the element to be grasped and/or manipulated and/or moved by the robot (10) by means of the predefined gripper (6), the elements being objects and/or people and/or places, the robot (10) being configured to interact with any element comprising at least one label (30) and said robot (10) comprising means for interpreting said identification and/or location and/or characteristic information and/or the ability of the element to be grasped and/or manipulated and/or moved by the robot (10) ), as well as means for determining behavior with respect to said element, in particular in terms of seizure and/or manipulation and/or movement. Robotic network architecture according to claim 1, characterized in that it further comprises a set of universal gripping means devices (1) configured to be graspable by means of the predefined gripper (6), the universal gripping means devices (1) being intended to be connected respectively to a set of objects of the set of elements, the universal gripping means devices (1) comprising at least one of said labels (30) and said universal gripping means devices (1) being configured to be grasped and/or manipulated and/or moved by means of the predefined gripper (6). Robotic network architecture according to any one of claims 1 to 2, characterized in that at least part of said labels (30) of said set of labels each comprise a capsule, a transmitting antenna and a battery, the transmitting antenna and the battery being housed in said capsule, and in that said robot (10) comprises a receiving antenna configured to communicate with said transmitting antenna. Network architecture according to claim 3 combined with claim 2, characterized in that said universal gripping means device (1) comprises a body and at least one fixing means, the body being configured to be grippable by the gripper (6 ) predefined and the fixing means being configured to connect the body to said object, said body extending mainly in a longitudinal direction between a first end and a second end, and the capsule being housed in a housing provided at one of said ends from the body. Network architecture according to claim 3 or 4, characterized in that at least part of said labels (30) further comprise at least one button disposed on an external face of the capsule and configured to activate/deactivate the label (30). ) and/or to call the robot (10) and/or to send a predefined message to the robot (10). Network architecture according to one of claims 3 to 5 combined with claim 2, characterized in that at least part of said labels (30) further comprise at least one inertial unit configured to detect a movement or an orientation of the elements who are equipped with it. Network architecture according to one of claims 3 to 6, characterized in that at least part of said labels (30) further comprise a Bluetooth chip coupled to the transmitting antenna. Network architecture according to claim 7, wherein the bluetooth chip meets standard 5.1 or higher and includes a direction finding function for determining an angle of arrival relative to an object also equipped with a bluetooth chip meeting the standard 5.1 or higher and includes a compatible direction finding function.