A feature of robots, compared to special machines, is the ability to program the robots.
The programming of robots can be a relatively simple or a highly complex operation. This depends upon the type of robot, the nature of the work it has to perform and its interaction with the environment and with peripheral devices. The task is compounded when more than one robot is working within the manufacturing area. A typical example would here be the automotive industry where robots are used for welding and spray painting car bodies
Two basic methods are used to program robots: on-line and off-line programming.
On-line programming is defined as the process of an operator directly interacting with the robot and the robot controller to teach the robot the new program. In on-line programming, the robot cannot be performing any other tasks while the operator is teaching the new program.
Conversely, in off-line programming, the robot program is developed on a computer which is not attached to the robot, so that robot can continue to perform an application. The following sections provide more detail concerning the processes of on-line and off-line programming.
Before proceeding, two terms must be defined: position and step. A position is a set of numbers (x,y,z) within a robot program which define the angles of each of the robot axes when the robot manipulator is at a point in space. The position is used by the robot controller to move each of the robot axes to the correct point during execution of the robot program.
A step is a command which is written in the robot programming language. The step causes the robot system to perform an action other than moving the manipulator. Examples of steps are: "close the gripper" and "wait for the operator to push the start button". (These example steps are written in English (or textual) for clarity-the actual steps would be written in the robot programming language.)
1) On-line programming methods
On line programming involves teaching the robot both the steps to perform and the positions to move to. In some robots, the positions are separate from the steps, in others the programs contain both the steps and the positions. Typically the steps are entered into the robot controller through some type of text editor. This is also known as the textual programming method. The positions can be also entered through the text editor, or can be "taught" by moving the robot physically to the desired location.
Two methods are used to teach positions: the walk-through, and the power-teach methods. In the walk-through method, the robot controller stores all of the positions of the robot as the operator moves the robot through the desired path. This method is used to store complex paths such as the paths used for spray painting. To move the robot, the operator moves a joy-stick, and the robot tracks the motion of the joy-stick.
* Continuous Path Method (Walk thru)
• Some applications (such as paint spraying) require the robot to follow a continuous and possibly complex path in space, rather than visit a sequence of points.
• In such cases a different type of robot is used which can be programmed by simply tracing out the required path in space.
• No record button is used but instead the computer automatically records joint positions at a rate of 80 to 100 times per second, as the robot is moving.
• In some cases the desired path is traced out using the actual robot. but this may be difficult due to the weight of the arm. Hence a number of manufacturers provide a second "teaching" arm, which is dimensionally identical but considerably lighter (it has no drives fitted) and therefore can be more easily manoeuvred by the programmer.
In the power-teach method, the operator uses a teach-pendant to move the robot to each position. Buttons on the teach pendant are pushed by the operator to cause single joints to move. Once the robot is at a position, the operator uses the teach pendant to store that position in the memory of the controller by pressing another button on the teach pendant. This method is used for pick-and-place operations, along with all other point-to-point moves of a robot.
* Point to Point Method (power-teach)
• A teach pendant, which in its simplest form is a box with buttons, is used to manoeuvre the robot in space
• The robot is driven to each of the points in space it must subsequently visit, in order to perform its intended task.
• At each point a "program" or "record" button is pressed and the current joint positions are automatically stored in the robot system's computer memory.
• When programming is complete, the program can be replayed using the stored sequence of joint positions.
• Variations in the above procedure include the positioning of the robot directly by hand, which is an option available with some robots.
• The buttons on a teach pendant may be used to move individual joints.
• However, when individual joints are moved, it may take some time to manoeuvre the end effector to the next point to be programmed.
• Faster programming can normally be achieved if the teach pendant's buttons can be used to drive the robot in directions parallel to the axes of a specified co-ordinate system
• The most obvious system is a cartesian system, where button presses move the robot in x, y and z directions.
• Some robots provide additional options such as cylindrical co-ordinates (r, e and z) or tool co-ordinates (which use a cartesian system permanently aligned with the end effector).
• In all cases the robot's computer uses a mathematical model of the arm to convert button presses requesting moves in the chosen co-ordinate system to the appropriate combination of joint moves.
* Other Program Information
• Other items of information must be added to the sequence of taught points in order to complete a robot program. These may include:
(a) The type of path to be followed between taught points
(b) The velocity, acceleration and deceleration of the arm along the replay path.
(c) Any loops or jumps to be included in the sequence of taught points during replay. (This may include conditions such as IF A GOTO B,where A is the status of a sensor signal sent to the robot and B is a numerically labelled taught point).
(d) Commands instructing the robot to sample an input signal or to generate an output signal.
As noted a teach pendant is normally used to generate program points but various methods are used to add the other items of information listed above, these include:
(a) A special purpose computer terminal.
(b) An "advanced" teach pendant which incorporates the front panel facilities described in (a) above.
(c) A general purpose computer terminal which is used in conjunction with a robot programming language. The language will be similar to a standard computer language, but will include commands to move to taught points and to specify the other program information described above.
* Choice of Replay Path
- Some robots will automatically move along straight lines between taught points during program replay, but many provide a choice of path, which may include:
(a) Joint interpolated control, where all joints simply start moving at the same time and stop moving at the same time. This produces a smooth path between taught points, but one which is not necessarily linear or easy to predict. Its advantage is that it requires the minimum computation and the robot may be able to move faster than would be possible if one of the following types of replay path were specified.
(b) Linear path control, where the arm moves in predictable straight lines between taught points.
(c) Circular path control, where the paths consist of circular arcs fitted through 3 successive taught points. (This option and (d) below are often used in arc welding applications, where the welding path is divided into a sequence of curves through taught points.)
(d) Any path which can be described mathematically. (Generally only available with robots which use a programming language).
* Input or Output Facilities
- In most robot cells, the robot will need to communicate with other items of equipment.
- Examples of inputs which could be received are a signal to indicate that a machine tool is ready to accept a component from the robot, a signal to tell the robot to unload a die casting machine or information as to the presence, location or identity of a component the robot is to acquire.
- Examples of outputs are signals to switch on a conveyor, a machine tool or a feeder.
Name:
The Real Programmer2013-04-21 23:45
Continue
2) Off-line programming
Three levels of off-line programming are in popular use:
l.basic, using a text editor;
2.intermediate, using a text editor and additional utilities (e.g. syntax checking programs); and
3.advanced, using graphic simulation systems, Computer Aided Robotics.
In the basic and intermediate off-line programming techniques, a text editor is used to define the steps and positions required. If the system which is used to develop the robot program includes the capability to check the syntax of the program, or determine if the robot can reach all of the positions, then this system can be classified as an intermediate system. Such intermediate systems are available from the robot vendors themselves at an additional cost. After the program is developed using either the basic or intermediate system, the program must be loaded onto the robot and tested before using it for production. The graphic simulation systems technique graphically portrays the programmed motion of the robot on a computer screen and can be quite helpful. Today when product life cycles are getting shorter and shorter it is very important to minimise the time from design to manufacturing. Case studies shows that an Off-Line Programming system would help to shorten this time considerably.
Other motivations for using an OLP system are amongst others:
• Frees the robot for productive use.
• Makes it possible for an engineer to evaluate equipment in a workcell and the layout of it. That is, it will assist engineers in making critical resource requisitions.
• When a proposed workcell has been designed, the entire application may be simulated in order to assess potential risk.
• Verifying the program, making preliminary confirmation of the cycle times.
• Such a system would work as a single reference point for multiple level decision-makers within a company. These can then collectively participate in the total workcell design.
Some of the most common off-line programming systems are:
• Workspace,
• CimStation,
• RobCad
• IGRIP etc.
IGRIP (Interactive Graphics Robot Instruction Program) is an interactive, 3D graphic simulation tool for designing, evaluating, and off-line programming robotic workcells. Actual robotic/device geometry, motion attributes, kinematics, clamps, fixtures, and I/O logic are incorporated to produce extremely accurate simulations.
The IGRIP 3D graphics-based simulation software package provides an interactive environment in which to verify production concepts, workcell designs, and manufacturing processes before implementing them on the shop floor. After verification is completed, automated factory floor devices, such as robots, can be programmed off-line based on the CAD data for the part being processed. IGRIP provides interactive program debugging for increased production reliability,
man/machine safety, and productivity.
The software includes built-in libraries of robots, common production equipment, and application-specific options that include tooling such as spot weld guns. These libraries, an integrated CAD system, and data translators for IGES, DXF, and other file formats support the rapid development of accurate simulation models. Simulation and analysis functions include automatic collision and near-miss detection, and automatic adjustment of a robot work envelope for tool offsets and joint limits.
Typically, an off-line simulator includes:
1. Robot Motion Planning
2. Cycle Time Analysis • Collision Detection
3. Multiple I/O Communication
4. Built-in Libraries of Robots and Related Equipment
5. Calibration and Signature Analysis to Automatically Simulate, Calibrate, Off-line Program, and Download Robot Programs from a Single Application.
6. Direct Translators for CATIA, Pro/ENGINEER, and UNIGRAPHICS
7. Stereo for Stereoscopic Depth Viewing Virtual Reality for Direct Interaction and Event Feedback within the Workcell
8. Calibration of robots
After a satisfying result has been given it is time to post-process the program and download it to the robot controller. But before this can be done we must be sure of that the Virtual models are correct according to the real ones. This leads us into the problems with CAR, namely the need to calibrate both the virtual workcell and robot. Because of the accuracy limitations of robots, any off-line programming can only approximate the robot motion and positions that will actually result. Accordingly, the positions programmed off line must be tested on the robot in actual working conditions before use in production.
A computer model of any real-time situation would always have some kind of inaccuracy in the model. Therefore a calibration is needed. Calibration can, as mentioned before, be divided into robot calibration and workcell calibration.
Robot calibration is needed due to:
• Differences between virtual and real machines, e.g. kinematics.
• Deviations due to manufacturing/mechanical deficiencies.
• Each robot kinematics changes over time due to wear and tear, service etc.
Since OLP relies on absolute accuracy and not repeater accuracy as in on-line programming we would get inaccuracies in the robot positioning (varies from one application to another). In some cases the majority of the robot work points has to be retaught with the actual robot to solve the problem. OLP systems will then lose their effectiveness.
There are different ways of calibrating a robot such as:
1. Manipulator signature calibration:
• Checking the kinematics, dynamics, gears, joints etc. Also gear lash, link lengths and orientations can be checked.
• This manipulator signature calibration may be needed once or twice a year.
2. Tool centre point offset calibration:
• Checking the tool length for wear.
• This calibration may be needed as often as every day.
3. Workcell calibration is needed due to:
• Difficulties in modelling an existing workcell (measuring etc.).
• Differences in shape of objects in virtual and real world.
When calibrating a workcell one checks the location of features such as: manipulators, fixtures, machine tools, conveyors etc. Another way is to just calibrate the fixture and the part.
There are many different calibration methods on the market with varying technology levels and complexity, but in general a calibration method should identify exact kinematic-geometric parameters, determine the elasticity and gear parameters and identify the geometric properties of the workcell (positions etc.). Some of the commercial calibration methods on the market are Robotrak and Optotrac.
Some questions :
Question 1: What is the power teach method ?
Question 2: Calibration can be divided into, what?
Question 3: Which methods can be used to calibrate a robot?