ITPD20110141A1 - LEARNING SYSTEM FOR ROBOTS FOR PAINTING WORKS - Google Patents

Description

TITLE

LEARNING SYSTEM FOR ROBOTS FOR WORKS OF

PAINTING

DESCRIPTION

The present patent relates to painting operations with anthropomorphic robots and in particular it concerns a new learning system for robots for painting surfaces in general with an emulation guide system.

Anthropomorphic robots are known, ie machines capable of automatically performing various types of operations in the most diverse applications. In order to perform an operation, these machines must be properly trained.

There are various ways to train robots, more or less costly from an operational point of view. It is clear that, once the robot has acquired an operation, the instruction cost per machined piece will be lower the greater the number of times the operation is repeated.

A typical use of anthropomorphic robots is known, namely that in the field of painting, which consists in the application, on surfaces of any kind, of paints, glues or other similar products in liquid or powder form. For example, the painting of surfaces in general, made of materials such as wood, metals, plastic, for example windows, furniture, chairs, tables, etc. is known.

In some cases, however, the machine instruction operation can be very onerous because in this type of applications the number of identical pieces to be treated is often quite small.

It is therefore advisable to have robot movement instruction tools that are quick and not particularly burdensome for operators.

Furthermore, for some types of objects to be painted in known robotic painting stations, the trajectories to be learned are often very complex and such as to require very long programming times. In paint stations, the methods currently available to train robots are as follows.

1. Point-to-point programming: consists of identifying and entering from the keyboard a discrete number of points along the ideal trajectories to be followed by the machine. The robot control automatically interpolates between the various points. It is a long, laborious process with an uncertain result, in fact it is difficult to find the ideal points and the trajectories drawn up by the machine are not always the best ones for carrying out the process.

2. Guide of the robot from keyboard by means of cursors or joysticks and memorization of movements: the system is laborious, difficult to execute and very imprecise, in fact it is not easy to make the machine execute the ideal trajectories.

3. Direct guidance of the robot using sensors to be applied to the robot itself: the operator acts on the sensors that send signals to the machine control, which activates the axes to make the robot's wrist move as desired and records its movements. This driving system is not simple, the trajectories are unnatural and the speed is very low for safety reasons.

4. Vision systems of the pieces with software for the automatic processing of the painting trajectories: they are widely used for the painting of predominantly one-dimensional or two-dimensional geometry artifacts. In the case of three-dimensional objects, however, the complexity, and therefore the cost, of the software would be such as to make this path practically impracticable.

5. Self-learning robots with direct guidance: these are very light and balanced machines, built in such a way that the operator can easily move the entire structure when the motors are not activated. It is therefore possible to have the robot simulate the movements to be performed during the painting cycle and memorize them for subsequent automatic repetition. The limits of this system are given by the dimensions of the machines, which, for obvious reasons, cannot be particularly large and / or robust; in addition, the painting movement is however a little hindered compared to manual painting by the fact of having to move the whole machine by hand. Finally, it should be noted that the robots must be made ad hoc for this type of application, i.e. it is not possible to use generic machines available on the market.

To overcome the limitations of the existing systems listed above, a new type of learning system for robots for painting surfaces in general with an emulation guide system has been studied and implemented.

The main aim of the present invention is to speed up, with a self-learning procedure, the instruction operations of one or more robots for painting surfaces in general, of any shape and material.

Another object of the present invention is to ensure a high quality of the coating result, because the learning is based on a learning coating, real or simulated, performed directly and manually by an operator, in the most natural way possible.

Another object of the present invention is to be suitable for teaching any type and number of robots for painting.

Another object of the present invention is to acquire the position and orientation data of one or more guns or painting devices that can be handled by the operator, in order to memorize any painting cycle, in real time.

Another object of the present invention is to memorize the painting cycles, then being able to repeat them identical for indefinite times, or to process and manage them to adapt them to specific needs.

The new learning system with emulation guide system provides that the operator, using for example a normal painting device, such as a gun, and manually maneuvering it in the simplest and most natural way possible, simulates or actually carries out the painting cycle, while the robot remotely receives the work done and memorizes it to repeat it indefinitely.

This guiding system comprises a detection instrument capable of perceiving the positions subsequently occupied by the gun, its orientation and, consequently, its movements carried out in manual painting and transmitting them to the control of a robot.

Said detection instrument makes use of one or more accelerometer and / or magnetic and / or optical and / or radiofrequency sensors or detectors and / or other devices useful for the purpose, designed to provide a measurement of the position and orientation of the gun with respect to a note theme.

System description

The new learning system includes:

· A learning station, equipped with at least one support for one or more pieces or objects or surfaces in general to be painted;

• at least one device or gun for manual painting or other device capable of being manipulated or used by an operator to perform all or part of one or more learning paints, real or simulated,

• a system for detecting one or more sequences of position and orientation coordinates in the space of said painting device, comprising for example at least one sensor positioned on the device itself;

· At least one control and data processing unit of said detection system, capable of communicating with the control panel of one or more robots capable of being instructed to perform the painting;

Said control unit, therefore, acquires the detected coordinate sequence, processes them transforming them into a sequence of movements for painting robots and communicates with said control panel of one or more robots, for their functionality.

The new learning system can be integrated or not in a painting station, also comprising one or more of said robots for painting as well as accessory components, such as a booth with fume extraction and possible conveyor of the articles to be painted.

In this case, said piece-holder support can be one of the supports or racks mounted on an automatic piece conveyor.

The new learning system also preferably comprises at least one switch or "micro switch" connected to the trigger of the painting device, for recording the start and end spraying phases. The new system also preferably comprises at least one portable keypad for returning the commands of various types of the equipment of the new station.

Said detection device comprises, for example, a signal transmitter, capable of measuring the position and orientation of the gun or device used for learning.

It can also be envisaged that said control panel of the robot also communicates with an automatic conveyor for the pieces to be painted.

System operation:

1. The operator holds the painting device, real or simulated, and controls the synchronization with the robot. In this phase it is possible to command that the robot already moves during the instruction or "teaching" phase or not.

2. Learning phase: the operator gives the start recording command and performs the simulated or real painting operation on a piece. The system has a whole series of controls that warn the operator if his movements go beyond the kinematic possibilities of the machine (position, speed, acceleration, curvature limits, etc.) and stops the acquisition.

3. At the end of the registration, the gun is positioned in the holster or support present in the painting station and the painting cycle just performed is stored in said control and processing unit;

4. At this point it is possible to immediately start Γ execution of the cycle, ie the automatic painting of the pieces, or proceed with the test and editing of the stored cycle. For example, in this phase it can be decided to redo one or more parts of the cycle, for example repaint a portion of the surface or one side of the piece, to increase or decrease the execution speed or to vary the distance of the gun from the piece.

5. Once a paint cycle is stored, it can be recalled when needed. It is also possible to select a number of pieces to be painted, at the end of which the robot will stop automatically. According to a possible embodiment, the pieces to be painted can be conveyed with an automatic conveyor which can consist of racks to one or more stations and, during the execution phase, the robot can be programmed to perform one or more identical pieces per rack.

It can be envisaged that said automatic conveyor may comprise a motor for the automatic rotation of the pieces to be painted.

For example, each rack can house two or more identical pieces. In this case, in the learning phase, only one piece can be painted, while the painting will be performed automatically on the others.

The new learning system can serve an unlimited number of robots.

It is also possible to save the painting cycle or program in several parts or "part-program", for example one for each side of the object to be painted.

Said "part-programs" saved can be launched according to a sequence defined by the operator, also combining them according to specific needs, for example by modifying the execution speed or by re-memorizing the trajectory of one or more of said "pari programs" without having to modify the others.

Fig. 1 shows a diagram of the new learning system (1) and the modalities of interconnections and communication between the components. The new learning system (1) comprises a learning station or cell (2), in which a support (3) is mounted for one or more pieces or objects or surfaces in general to be painted, said support (3) being preferably movable to rotate the workpiece, for example manually or automatically.

The new system also comprises at least one device, for example a gun (6), for manual painting, or other device suitable for use or manipulation by an operator, which performs the first learning painting, simulated or real, on at least one piece or part of it, performing the operation in the most natural way possible.

The new system also includes a detection system (7) of the position and orientation in space of said painting device (6).

Said detection system (7) can for example be of the type with accelerometric and / or magnetic and / or optical and / or radiofrequency detectors or other devices, and is able to detect position and orientation measurements of said painting device (6) , with respect to a known theme.

The new system also comprises at least one unit (8) for controlling and processing the data acquired by said detection system (7), also able to communicate, directly or indirectly, with at least one control panel (5) of at least one robot (4) for painting. It can be envisaged that said control panel (5) also controls an automatic conveyor (10) for the pieces to be painted.

The new learning system also preferably comprises at least one keypad (9), preferably portable, for sending back commands of various types, said keypad (9) being connected to said control unit (8).

These are the schematic modalities sufficient for the skilled person to realize the invention, consequently, in concrete application there may be variations without prejudice to the substance of the innovative concept.

Therefore, with reference to the preceding description and to the attached table, the following claims are expressed.

Claims (9)

CLAIMS 1. Learning system (1) for robots (4), for painting operations on surfaces in general, characterized by the fact of comprising: • a learning station or cell (2); • a support (3) for one or more pieces or objects or surfaces in general to be painted; • at least one device (6) for manual painting installed in said station (2) and able to be used or manipulated manually by an operator to perform at least a first learning painting, real or simulated, on at least one piece or part of it; • at least one system for detecting (7) one or more sequences of position and orientation coordinates in the space of said painting device (6) during said learning painting; • at least one control and data processing unit (8), able to acquire and process said one or more sequences of coordinates detected and to communicate with at least one control panel (5) of one or more robots (4), for the functionality of the same in the repetition of all or part of said sequences of acquired coordinates. 2. Learning system (1) for robots (4), as per claim 1, characterized in that it comprises one or more of said robots (4) for painting with relative at least one control panel (5), and where one or more of said robots (4) repeat said sequences of coordinates acquired also in real time, during said learning phase, simultaneously with the movement of said painting device (6). 3. Learning system (1) for robots (4), as per claims 1, 2, characterized in that, during said first learning painting, said control unit (8) performs compatibility checks of said acquired coordinate sequence with the kinematic and dynamic possibilities of said painting robot (4). 4. Learning system (1) for robots (4), as per claims 1, 2 3, characterized in that said control unit (8) stores said sequence (s) of coordinates in one or more parts, and where each part can be individually processed, repeated by said robot (4) or modified independently from the others. 5. Learning system (1) for robots (4), as per claims 1, 2, 3, 4, characterized in that said piece-holder support (3) is movable, manually or automatically. 6. Learning system (1) for robots (4), as per the preceding claims, characterized in that said detection system (7) is based on accelerometric sensors or detectors and / or magnetic sensors or detectors and / or optical sensors or detectors and / or with radiofrequency sensors or detectors and / or including other detection devices. 7. Learning system (1) for robots (4), as per the preceding claims, characterized by the fact of comprising at least one keypad (9), preferably portable, for sending back commands of various types, said keypad (9) being connected to said control unit (8). 8. Learning system (1) for robots (4), as per the preceding claims, characterized in that said painting device (6) comprises a switch connected to said control unit (8), for recording the start phases and fine spraying. 9. Learning system (1) for robots (4), as per the preceding claims, characterized in that said control panel (5) of said one or more robots (4) is connected to at least one external automatic conveyor (10) of pieces to be painted, said conveyor (10) in turn comprising one or more racks at one or more stations.