FR3044573A1 - METHOD AND INSTALLATION FOR CONSTITUTING A BATCH OF PARTS FROM PARTS LOCATED IN DIFFERENT STORAGE AREAS - Google Patents

Abstract

Method of automatically constituting, by means of a polyarticulated system (7) secured to at least one vision system (2), at least one batch of pieces (6) in a receiving zone (10) from of parts located in at least two different storage areas (5, 13), said polyarticulate system comprising at least two separate gripping members, each capable of gripping a part (6) by at least one specific gripping area. According to the invention, the method comprises at least the steps of: - automatically moving the polyarticulate system (7) to a first storage area (5, 13) and: (i) capturing an image of the storage area ( 5, 13) by means of the vision system (2); (ii) processing the information resulting from the image and identifying all the gripping zones of the parts (6) compatible with the gripping members; (iii) locate, in position and orientation, the identified catchment areas; (iv) choosing the gripping zone that has the best chance of being grasped by a gripping member and which presents no collision stress, and automatically calculating, for said corresponding gripping member, a gripping trajectory of the piece by the chosen catch area; (v) gripping the corresponding piece (6) according to the calculated trajectory, and automatically moving the polyarticulate system (7) to the receiving area (10) to deposit said piece (6) therein; - Move automatically the polyarticulate system (7) to a second storage area (5, 13) and perform steps (i) to (v) again.

Description

METHOD AND INSTALLATION FOR CONSTITUTING A BATCH OF PARTS FROM PARTS LOCATED IN DIFFERENT STORAGE AREAS

TECHNICAL AREA

The present invention relates to the technical sector of handling parts by a polyarticulate system. By the term "polyarticulate system" is meant a six-axis robotic system for example. The invention more particularly relates to a method for automatically constituting, and via a polyarticulate system subject to a vision system, a batch of separate parts in a receiving area (dispensing area) from separate parts located in at least two different storage areas (sampling areas). The invention finds an advantageous application in the so-called "Kitting" (Anglo-Saxon) processes, in which separate but related individual pieces are grouped together to form a kit, ie a batch or a collection. the rooms. The said pieces can then be assembled and / or packaged and / or delivered together.

More generally, the invention finds an advantageous application in the so-called "picking" (Anglo-Saxon) methods, consisting of taking one by one parts for the preparation of orders for example.

In the applications concerned, the storage areas can be of different types without departing from the scope of the invention. They may consist, for example, in fixed areas receiving bulk amassed parts, or consist of moving areas such as conveyors conveying said bulk parts. By the term "bulk" is meant a cluster of pieces randomly oriented in a volume or flat. Each storage area comprises a plurality of identical pieces.

Prior Art Generally, in the state of the art, the so-called "picking" or "kitting" operations are either carried out manually by an operator, or carried out by a polyarticulated system subject to a vision system for seize and deposit coins.

However, the type of polyarticulate system used is of the type requiring learning of the part to be gripped and a step of recognition of the part.

A polyarticulate system of this type is in particular known from the patent application FR 2 987 685, which describes a method of controlling a robot for moving at least one part arranged on a support in a 3D space. This method notably comprises the steps of: acquisition of at least two 2D images of the 3D space; segmentation of each previously acquired 2D image, intended to extract from each 2D image at least one 2D geometric primitive; - search for matches between 2D geometric primitives extracted from distinct 2D images; calculating the coordinates of a 3D geometric primitive contained in the 3D space for each group of corresponding 2D geometric primitives; associating each 3D geometric primitive with a known piece; searching for the position and the orientation in the 3D space of each known piece to which at least one 3D geometric primitive has been associated; - Movement via a robot of at least one of the known parts.

Thus, this method offers the possibility of detecting, locating, and orienting parts in a 3D space.

However, as mentioned, the method has the disadvantage that it does not adapt to any type of part to move. Indeed, this method works only with previously known parts and identified by a learning software. This method does not make it possible to introduce, during the process, an unknown part reference, or even to change the constitution of a batch. Indeed, to change the constitution of a batch, it is necessary to load another suitable program for controlling the polyarticulate system. SUMMARY OF THE INVENTION The invention aims to overcome the aforementioned drawbacks, and thus to provide a method and an installation for automatically constituting a batch of parts, without prior learning of said parts, and without prior learning of the trajectories of this. For this purpose, a method has been developed for automatically constituting, by means of a polyarticulate system subjected to at least one vision system, at least one batch of pieces in a reception zone from parts located in at least two different storage areas, said polyarticulate system comprising at least two separate gripping members, each capable of gripping a part by at least one specific gripping area.

According to the invention, the method comprises at least the steps of: - automatically moving the polyarticulate system to a first storage area and (i) capturing an image of the storage area by means of the vision system; (ii) processing the information resulting from the image and identifying all gripping areas of the parts compatible with the gripping members; (iii) locate, in position and orientation, the identified catchment areas; (iv) choosing the gripping zone that has the greatest chance of being gripped by a gripping member and which presents no collision constraint between said gripping member and the environment, and automatically calculating, for said gripping member corresponding, a trajectory of taking the piece by the chosen catch area; (v) enter the corresponding piece according to the calculated trajectory, and automatically move the polyarticulate system to the receiving area to deposit said piece; - Move automatically the polyarticulate system to a second storage area and perform steps (i) to (v) again.

In this way, the polyarticulate system decides for itself and in real time the best way to take a part. The polyarticulate system moves or is moved from zone to zone to consecutively take parts with very different geometry and thus constitute a batch of parts.

The polyarticulate system operates without learning so that the introduction of new part references is done without difficulty and without constraint. The polyarticulate system does not seek to recognize the part to be grasped, but seeks to recognize specific gripping zones on said part to be gripped. The part can be arbitrary, it is only necessary that a gripping area compatible with a gripping member is detected so that the part is grasped by this gripping zone.

Since the gripping zone is located in position and orientation, the method requires no learning on the gripping trajectory of the corresponding coin. The capture trajectory is calculated in real time, and for each piece to capture. The method then makes it possible to perform operations of very rapid take and deposit of unknown parts. The advantage of the invention lies in the fact that the polyarticulate system is autonomous. After having captured and processed an image, he finds and takes into account all the possible gripping zones of the parts to be gripped, compatible with his gripping organs, he takes into account all the gripping devices available and puts them in competition, he takes it takes into account all the collision constraints of the gripping members for a given input that can interfere with said input, and finally it decides on the best gripping area to be gripped and the best gripping member to be used.

After gripping a part, the multi-articulated system moves or is moved to the receiving area to deposit the workpiece and moves or is moved to another storage area to enter another component part of the batch that it is in. train to prepare. Advantageously, and in the case where the storage areas comprise parts storage bins, the method comprises the steps of detecting an empty storage bin, grasping it and evacuating it.

In the case where the storage bins comprise packaging components, such as spacers, braces, etc., the method comprises steps of recognizing said packaging components, seize and evacuate.

Preferably, when an empty storage bin is evacuated, a storage bin comprising parts to be input is automatically reintroduced into the storage area. Advantageously, and in the case where the receiving zone comprises parts receiving bins, the method comprises the steps of detecting a full receiving tray, grasping it and evacuating it.

Preferably, when a full receiving bin is evacuated, an empty receiving tray is automatically reintroduced into the receiving area.

In this way we speak of adaptive robotics, intelligent and autonomous.

According to a particular embodiment, the step of identifying compatible gripping areas consists in identifying continuous and flat areas in the case where the gripping member is of the suction cup type.

By the term "suction cup" is meant both a suction cup operating with an air suction, a sucker magnetic type such a magnet for example. It is thus obvious that to be compatible with the suction cup, the identified continuous region must have a continuous region comprising a minimum surface corresponding to the active surface of the suction cup.

According to another particular embodiment, the step of identifying specific compatible areas consists of identifying edges, generatrices or parallel surfaces in the case where the gripping member is of the clamp type, the maximum distance between said edges, generators or surfaces to be compatible with the kinematic characteristics of said clamp. For example, the maximum deviation must be less than the maximum possible gap between jaws of said clamp. The invention also relates to an installation for the automatic constitution of at least one batch of parts in a reception zone from rooms located in at least two different storage areas, said installation comprising a polyarticulate system subjected to at least one system of vision capable of capturing an image of the storage areas, and comprising at least two distinct gripping members, each able to grip a part by at least one specific gripping zone.

According to the invention, said polyarticulate system comprises means for moving between the storage areas and the receiving zone, and the vision system and said polyarticulate system are subject to processing and calculation means capable of processing the information. resulting from the captured image to identify all gripping areas compatible with the grippers, to locate, in position and orientation, the identified catch areas, to select the catch area that has the greatest chance of being gripping by a gripping member and having no collision constraint between said gripping member and the environment, for automatically calculating, for the corresponding gripping member, a gripping trajectory of the piece by the chosen gripping zone, and to control the movement of the polyarticulate system between the storage areas and the receiving area to deposit the input part.

This installation is thus able to implement the method according to the invention and comprises in this way all the aforementioned advantages of said method.

According to particular embodiments, the vision system is movably mounted, for example on the polyarticulate system or on rails, between the storage areas. The vision system can also be mobile between the storage and receiving areas. According to another example of implementation, the installation comprises several vision systems, each fixed above the storage areas, and preferably also above the reception area.

BRIEF DESCRIPTION OF THE DRAWINGS Other characteristics and advantages of the invention will emerge clearly from the description which is given below, by way of indication and in no way limiting, with reference to the appended figures, in which: FIG. 1 is a diagrammatic representation perspective illustrating a first example of implementation of an installation of automatic constitution of a batch of parts according to the invention; FIG. 2 is a schematic representation in perspective illustrating a second example of implementation of the installation according to the invention.

DETAILED DESCRIPTION OF THE INVENTION The invention relates to a method and an installation for constituting automatically, and via a polyarticulate system subject to a vision system, at least one batch of pieces in a reception zone from parts in at least two different storage areas, especially for "picking" and / or "kitting" applications. The installation comprises a polyarticulated system, at least one receiving zone of the batch of parts, and at least two different coin storage areas for the constitution of said batch.

The polyarticulate system is, for example, a six-axis robot well known in the state of the art. Said polyarticulate system is able to grip a part by a specific gripping zone of said part. For this purpose, said polyarticulate system comprises at least two gripping members, and preferably three, and is subject to at least one vision system.

The gripping members may be of any suitable type. The bottom line is that they are able to grab a piece by a specific gripping area. They are, for example, in the form of a clamp comprises two jaws able to move away from each other and to grab a room. The specific grip areas of a workpiece that can be grasped by a clamp are edges, generatrices or surfaces, preferably parallel. Obviously, the maximum distance between said ridges, generatrices or surfaces must be compatible with the kinematic characteristics of said clamp. For example, the maximum gap must be less than the maximum possible gap between the jaws of said clamp.

The gripping members may be of suction or magnetic sucker type, comprises, for example, at its end a resilient member in the form of a sleeve adapted to be applied to a specific area of a part for grasping by suction or magnetism said room. The specific areas of a room that can be gripped by a suction cup are continuous regions, namely substantially flat surfaces and dimensions adapted to receive bearing the elastic sleeve of the suction cup.

The coins to be seized are received randomly and in bulk in the storage areas. The storage areas can thus include storage bins of parts, possibly with packaging components inside the bins, such as dividers, braces, etc ... for the arrangement and separation of parts to inside the bins. The storage areas can be fixed, or mobile by the implementation of conveyors. The pieces are for example of sizes, shapes, and natures different from one storage area to another.

According to the invention, the polyarticulate system is movable between the storage areas and the receiving area of the batch of parts. The moving means are of all appropriate types. For example, the multi-articulated system can be autonomous in its movement and move itself, and / or be subject to distinct displacement means, such as for example mounted on rails or on board a trolley or a rolling vehicle, driven automatically. . The polyarticulate system can also be fixed with an arm movable from one area to another.

The vision system is of any suitable type, such as for example a 3D camera, or a camera in stereo mode. The vision system can be arranged directly on the polyarticulate system so as to be movable with it, from zone to zone, to capture images of said zones. According to other exemplary embodiments, the vision system is movably mounted on rails between the storage areas, and advantageously between the storage and reception areas. In another example, the installation comprises several vision systems, each fixed above the storage and reception areas.

In addition, the installation comprises processing and calculation means subject to the polyarticulate system and the (s) system (s) vision.

These processing and calculation means are of any appropriate type, and may for example be in the form of a microprocessor integrating a processing and calculation software.

These processing and calculation means are able to process the information resulting from the captured image so as to identify, on the parts to be gripped, gripping zones compatible with the gripping members, such as, for example, edges, generatrices, or surfaces preferably parallel, whose distance is compatible with the kinematic characteristics of the jaws of a clamp, or flat surfaces whose dimensions are adapted to receive in support of an elastic sleeve of a suction cup.

Said processing and calculation means also make it possible to locate, in position and in orientation, said identified compatible specific gripping zones.

The processing and calculation means are suitable, when the gripping zones are located, to automatically choose the gripping zone having the best chance of being correctly grasped by one of the gripping members. The different gripping organs are therefore put in competition.

The choice of the gripping zone that does not have the better chance of being correctly grasped consists, for example, in weighting each identified gripping zone with a coefficient which is a function of the success probabilities of gripping by the gripping member. corresponding, and to choose the area with the most probabilities of successful entry. The gripping members are therefore put in competition and it is the one with the most chance of success that is activated to enter the piece. In practice, it is a question of assigning weighting coefficients for each identified and compatible catch area, as a function of one or more parameters. The parameters used can be of any kind such as, for example, the altitude of the zone to be grasped, the area of the visible surface, the inclination of a gripping zone. Each catch zone therefore comprises a weighting for each of the chosen parameters. Then an average of each weighting of the shooting zone is made to obtain a score. The catch zone having the largest score, or the smallest depending on the nature of the weighting, is chosen to be captured.

According to another characteristic of the invention, the processing and calculation means are also able to identify, for the selected engagement zone, the collision stresses between the gripping member able to grip the part by the selected zone and the 'environment. Indeed, depending on the disposition of the part whose gripping area has been chosen, the compatible member may have a gripping constraint. The part to be grasped may be, for example, too close to the edge of a storage tank or another room so that the gripping member will be obstructed and will not be able to grip the part without striking the edge of said tray or the adjacent room. Other collision constraints may occur.

In practice, the volume of space of the gripping member is known by the processing and calculation means, it is therefore necessary to check, from the captured image of the storage area, the intersection between the bulk volume of the gripping member and the environment around the chosen gripping zone. If there is a collision constraint, namely a risk of interference between the gripping member and the environment of the chosen catch area, said gripping zone is abandoned and the second gripping zone having the most chance to be seized is chosen. When no collision constraint is noted, the processing and calculation means automatically calculate, for said corresponding gripping member, a trailing path of the piece by the chosen gripping zone. The installation according to the invention thus makes it possible to implement an automatic constitution process, and without learning, at least a batch of pieces in a reception area from rooms located in at least two different storage areas.

According to the invention, the installation described above is capable of implementing the following succession of steps: automatically moving the polyarticulate system towards a first storage area and: (i) capturing an image of the zone of storage by means of the vision system; (ii) processing the information resulting from the image and identifying all gripping areas of the parts compatible with the gripping members; (iii) locate, in position and orientation, the identified catchment areas; (iv) selecting the gripping zone that is most likely to be grasped by a gripping member and that has no collision stress between said gripping member and the environment of the gripping zone, and automatically calculating, for said corresponding gripping member, a gripping trajectory of the piece by the chosen gripping zone; (v) enter the corresponding piece according to the calculated trajectory, and automatically move the polyarticulate system to the receiving area to deposit said piece; - Move automatically the polyarticulate system to a second storage area and perform steps (i) to (v) again.

This method is advantageous in that it allows the polyarticulate system to move from one area to another and to grab and deposit any parts, without learning the parts, nor the trajectories of these, to constitute a batch.

Referring to Figure 1, there is described an example of application of the method according to the invention. The installation (1) comprises a vision system mounted on a linear rail (3) and able to move along and to the right of two conveyor lines (4) receiving several different storage bins (5) of parts (6). ) in bulk. In the illustrated example, the conveyor (4) comprises for example two rows of three storage bins (5) each having different pieces (6), to form a batch of three pieces. The installation comprises a polyarticulated system (7) subject to the vision system (2) and embedded on a vehicle (8) movable on a rail (9) parallel to and adjacent to the conveyors (4). The vehicle (8) also loads bins (10) to receive lots of parts to constitute. Preferably, the receiving bins (10) are arranged on conveyors (11) adapted, by the displacement of the vehicle (8) to be positioned opposite an outlet (12) for full receiving bins or, for example, a conveyor for supplying the vehicle (8) in an empty receiving tray. It is also possible that the full or empty receiving bins (10) are evacuated or fed by an operator.

Thus, the vision system (2) moves on the rail (3) to capture images of the storage bins (5). Once the image is captured and processed, the polyarticulate system (7) is able to grab a part (6) and comes to deposit it in bulk in a receiving tray (10) whose coordinates it knows since it is on the same reference system as the polyarticulated system (7). A second vision system may be considered to determine the position and orientation of the receiving tray.

The polyarticulated system (7) repeats the capture and removal operation for the two other pieces (6) present in the two other storage bins (5).

Thus, the batch of pieces has been seized and deposited in a receiving tray (10), it can be evacuated and a new empty receiving tray (10) can be fed.

The vision system (2) recognizes the presence of packaging components, such as spacers or braces between the parts of the storage bin (5). These packaging components, previously stored by the processing and calculation means, are recognized, captured and removed by the polyarticulate system (7). In the same way, the vision system (2) detects the presence of empty storage tanks (5) and, after locating their position via the computing and processing means, the polyarticulate system (7) seizes them and evacuate them.

Referring to Figure 2 which illustrates a second installation (1) for implementing the method according to the invention, the polyarticulate system (7), subject to the vision system (2), seized one by one parts (6). ) located in different storage areas (13) to deposit them in a receiving tray (10) to form a batch. The vision system (2) is rotatably mounted about an axis to scan the different storage areas (13) arranged around the polyarticulate system (7). The receiving tray (10) full is then, for example, emptied into a bagger to package together the parts of the same batch.

In this embodiment, the displacement means of the polyarticulate system (7) are constituted by its arm which is able to move between the storage areas (13) and the receiving tray (10).

It is obvious that the method and the installation (1) according to the invention can implement a plurality of polyarticulated systems (7), each of which can be subject to the same vision system (2), or each may be subject to his own vision system. Thus, when one of the polyarticulate systems moves an input part, another polyarticulate system can, in masked time on the displacement of said input part, to enter another, and so on. This makes it possible to perform operations of taking and removal of unknown parts in an optimal and fast manner.

Claims (11)

claims A method of automatically constituting, by means of a polyarticulate system (7) secured to at least one vision system (2), at least one batch of parts (6) in a receiving zone (10) from pieces located in at least two different storage areas (5, 13), said polyarticulate system comprising at least two distinct gripping members, each capable of gripping a part (6) by at least one specific gripping zone, characterized in that it comprises at least the steps of: - automatically moving the polyarticulate system (7) to a first storage area (5, 13) and: (i) capturing an image of the storage area (5, 13) by means of the vision system (2); (ii) processing the information resulting from the image and identifying all the gripping zones of the parts (6) compatible with the gripping members; (iii) locate, in position and orientation, the identified catchment areas; (iv) choosing the gripping zone that has the greatest chance of being gripped by a gripping member and which presents no collision constraint between said gripping member and the environment, and automatically calculating, for said gripping member corresponding, a trajectory of taking the piece by the chosen catch area; (v) gripping the corresponding piece (6) according to the calculated trajectory, and automatically moving the polyarticulate system (7) to the receiving area (10) to deposit said piece (6) therein; - Move automatically the polyarticulate system (7) to a second storage area (5, 13) and perform steps (i) to (v) again. 2. Method according to claim 1, characterized in that the storage areas comprise storage bins (5) of the rooms, and in that the method comprises the steps of detecting an empty storage bin (5), on the seize and evacuate. 3. Method according to claim 2, characterized in that the storage bins (5) comprise packaging components, and in that the method comprises steps of recognizing said packaging components, seizing them and clear out. 4. Method according to claim 2, characterized in that when an empty storage tank (5) is evacuated, a storage bin (5) comprising parts (6) to enter is automatically reintroduced into the storage area. Method according to any one of claims 1 to 4, characterized in that the receiving area comprises receiving bins (10) of the pieces, and in that the method comprises steps of detecting a receiving tray ( 10) full, grab it and evacuate it. 6. Method according to claim 5, characterized in that when a solid receiving tray (10) is evacuated, an empty receiving tray (10) is automatically reintroduced into the receiving area. 7. Method according to any one of claims 1 to 6, characterized in that the step of identifying compatible gripping areas consists of identifying continuous and flat areas in the case where the gripping member is sucker type . 8. Method according to any one of claims 1 to 7, characterized in that the step of identifying compatible specific areas consists of identifying edges, generatrices or parallel surfaces in the case where the gripping member is clamp type, the maximum distance between said edges, generators or surfaces to be compatible with the kinematic characteristics of said clamp. 9. Installation (1) automatic constitution of at least one batch of pieces (6) in a receiving area (10) from pieces (6) located in at least two different storage areas (5, 13), said installation (1) comprising a polyarticulated system (7) subjected to at least one vision system (2) capable of capturing an image of the storage areas (5, 13), and comprising at least two distinct gripping members, each capable of gripping a part (6) with at least one specific gripping zone, characterized in that said polyarticulate system (7) comprises displacement means (8) between the storage areas (5, 13) and the receiving zone ( 10), and the vision system (2) and said polyarticulate system (7) are subject to processing and calculation means capable of processing the information resulting from the captured image to identify all the capture areas compatible with the gripping members, to locate, in position and in rientation, the gripping zones identified, to choose the gripping zone that has the best chance of being grasped by a gripping member and that presents no collision constraint between said gripping member and the environment, to automatically calculate for the corresponding gripping member, a gripping path of the workpiece (6) by the selected gripping zone, and for controlling the movement of the polyarticulate system (7) between the storage zones (5, 13) and the zone receiving (10) to deposit the piece (6) entered. 10. Installation according to claim 9, characterized in that the vision system (2) is movably mounted between the storage areas (5, 13). 11. Installation according to claim 9, characterized in that it comprises several vision systems (2), each fixed above the storage areas (5, 13).