"Almost anything in software can be implemented, sold, and even used given enough determination. There is nothing a mere scientist can say that will stand against the flood of a hundred million dollars. But there is one quality that cannot be purchased in this way - and that is reliability. The price of reliability is the pursuit of the utmost simplicity." - from 'The Emperor's Old Clothes' by C.A.R.Hoare

  • ROBOFORTH II is the software that runs in the robot controller. It's a fourth generation language. That means it uses simple natural language to describe what you want to do.
  • It's interactive. You can type a command and get immediate movement from the robot, or action on the I/O such as valves or instruments.
  • It's human - using English language words, which the user can use like building blocks to define more powerful words. Structured conditionals allow decision-making branches and loops.
  • It has extensive I/O to control and interface with external devices, synchronized with robot motion if required.
  • Easy positional programming (teaching) - named positions and paths, relative and absolute positions, Cartesian or joint control, palletizing, continuous path and a simple teachpad.
  • RobWin is the computer GUI project management software under Windows that brings all the features of ST robots together on one screen in a 'tethered' programming environment.
  • Supervisors include RobX, ST's ActiveX wrapper for easy integration into larger systems for use with VB, C, LabVIEW and other Windows based systems.
  • 24/7 technical support by email or telephone.

The most extensive and powerful robot control system on the planet

ROBOFORTH II is a 4GL robot control language designed to cover every eventuality in programming your robot, whether it be coating, assembly, laboratory handling, testing, or whatever. There are various means of acquiring spatial data and using it such as matrices, continuous path, object tracking, collision avoidance, plus numerous input-output features making it easy to interface with and control peripheral equipment at the same time as control the robot arm. Above all Roboforth is a real language i.e. has an extensive vocabulary of words, which, like a natural language, help you express your ideas to the system. As a language Roboforth is a means of communication between the human user and the robot arm: for the user to tell the robot what to do and how to do it; for the robot to tell the user what it is doing and what it knows.

Forth and Roboforth are organized as a linked list of words known as the dictionary. Together with each word in the dictionary is its definition, which is usually in terms of words lower down in the dictionary except for 'primitives' at the bottom, which are native machine language. There are around 400 words in Roboforth available to the user. Programming in Forth and Roboforth consists of creating new words into the dictionary whose definitions are in terms of words already defined. New words may include data e.g. positional information about the robot. Values (or arguments) are passed between words on a stack.

Any given position of the arm may be specified as coordinates relative to a central position for the arm known as the HOME position. In ROBOFORTH II these coordinates are in terms of the movement required by the motors of the arm to reach that position i.e. number of motor 'bits' counted from the HOME position. Alternatively they may be expressed in millimeters in Cartesian X,Y,Z coordinates. To teach the robot you first move the arm to the required position using any of the means provided and secondly record the coordinates of that position. This record has a header which is another entry in the dictionary. When replayed the arm moves from one position to the next in turn (point-to-point programming). Each position is a list of values plus flags to indicate what kind of data it is e.g. Cartesian coordinates, absolute or relative coordinates, name of any object located at that position etc. The robot can also learn positions itself, for example by searching or by computing a target position or by modifying a taught path to suit a new requirement.

In ROBOFORTH II there are various means of moving the robot and various ways of teaching the robot. These are distinctly separate. In practice the procedure is to first move the robot to the desired position then learn it. Positions or lists of positions have names, which are provided by you, the user and the names or 'words' form headers. Finally the words are used in procedures that are also entries in the dictionary. These determine which positions the robot goes to and when/why.

The gripper may also be operated or told when to operate. Similarly any activity may be tested in immediate mode or 'compiled' into the definition of a new word. Error conditions can be re-programmed so, for example you can determine what the robot does and what happens to the associated equipment in the event of the stop button being pressed or in the event of a stall because of jammed product. These and other features make ROBOFORTH II the most powerful robot control system in existence.

While Roboforth runs in the robot controller its counterpart, RobWin runs in the computer. RobWin is only used for programming and may be dropped once the robot is programmed.

RobWin provides a communications window through to the controller but also allows the user to quickly access many of the features of Roboforth by clicking buttons and by completing 'dialog boxes' and maintains disk files.

Moving the robot and teaching the robot are two separate operations.

Moving the robot
may be achieved with any or a combination of four methods:
1. Using a teach pad:
After clicking the teach button on screen simply select a joint or axis on the teach pad then press the + key to move in one direction or - in reverse; movement continues as long as the key is pressed. Low speeds are appropriate for accurate positioning and speed may be changed up or down on screen or from the teach pad.
2. Commanding each joint:
Commands such as TELL SHOULDER 1000 MOVE which moves the shoulder axis 1000 counts. Other commands are: MOVETO (an absolute command), GRIP, UNGRIP, HOME, CARTESIAN, JOINT, ALIGN/NONALIGN and so on. There are about 250 similar commands and settings in Roboforth and a total vocabulary of around 800 words.
3. Positioning in Cartesian coordinates: (reverse kinematics)
Dialog boxes permit positioning in absolute or relative coordinates. Alternatively MOVE or MOVETO commands may be entered or included in a program. Wrist/hand angles may be specified or controlled. Tool transformations depend on robot model and number of axes.
4. Jog.
The teach pad may be used in 'Jog' mode in which the keys represent X, Y, Z, pitch and roll axes. Each press of the + or - key causes the robot to move a set increment in the selected direction. The increment may be reduced on screen or from the teach pad as the robot nears the target position.
5. 'Lead-by-the-nose'
As with most robots it is possible to de-energize the robot, position it by hand and learn that position. However the skill required to position a robot to better than a millimeter is beyond most people so we don't recommend it.

Teaching the robot
may be achieved by choosing one of two 'entities':
1. A Route:

This is a list of spatial coordinates each of which is a row of numbers representing motors counts away from the zero (HOME) position or absolute Cartesian coordinates. The route is created by the user with his/her given name e.g. ROUTE I66. RobWin makes this simple with a dialog box that creates the entry in the controller as well as on disk. Coordinates are then added to the list by clicking 'insert position' or using the tick key on the teach pad. The robot moves from point to point with the command RUN - either stopping at each point or moving through them. Associated commands provide editing and the ability to run parts of a route, or to retrace. A route is also used as a reference for discrete positions, for palletizing for example. Editing is achieved using dialog windows or with ROBOFORTH commands such as REPLACE which also permit self learning features, for example the robot can modify its own positions according to the programmed procedure. Functions such as gripper operation, delays, speed changes, spray/glue on/off etc. can be embedded in the route to take effect in the required sequence.
A Matrix is also a list, organized in 2 dimensions and a row is one-dimensional. The number of rows and columns are specified in dialog boxes, the corners of the matrix learned and the system computes the rest of the positions and downloads them to the controller.
2. A Place:
This is a single named coordinate. It is self learning and self executing. It is created by the user with his/her given name using a dialog box or with a native command e.g. PLACE JIG.
To return the robot to this position later simply use the word JIG.

Finally all these learned and named entities are used in the procedure file to create new definitions or 'words' which determine how the positions are used, i.e. in what order, in what circumstances etc. For example a word might be defined using a PLACE named JIG and a matrix route named TRAY:

: GETPART
TRAY INTO
GRIP
UP
JIG
UNGRIP
WITHDRAW
;

Because Forth and Roboforth are stack oriented the position within the TRAY is given on the stack just before the command, for example
5 GETPART will get a part from position 5 of the TRAY and put into the JIG
This may be typed as a command into the communications window, or sent from a supervisor, which computes the position number, or further compiled into an even higher level definition.


Extensive input-output is also provided, which is easily programmed and integrated with robot motion, for example suppose a pump were connected to parallel port PA on bit 5. You could add a definition thus: 

: PUMP PA 5 ;

You can then control the pump with simple PUMP ON and PUMP OFF. This could then be included in your higher level definition: 

: GETPART
TRAY INTO
GRIP
UP
JIG
UNGRIP
WITHDRAW
PUMP ON
5000 MSECS
PUMP OFF
;
Advanced features

Include the ability to re-program errors such as what happens when you press the stop button. The robot always stops with controlled deceleration when the 'soft' stop is pressed, and normally the system aborts. However it is possible to program alternative action using 'vectored execution', for example turn on a light and wait for the green button to be pressed, then perform some recovery procedure and continue with the main task. The result is rather like an interrupt, which is also programmable as are positional errors. For example if there is a jam in the machine with which the robot is working the robot may crash into it, causing a positional error, which normally aborts the system. This could be reprogrammed to pull back, flash a light then wait for an operator to clear the condition and press the continue button. Suppose the robot is carrying a rotary cutter. When the stop button is pressed the cutter must stop also. When the continue button is pressed the robot can continue; the cutter is turned back on but only if it was on when the stop was pressed.

Object database to track what objects are in what positions.

Sensors: The dual processor arrangements permits one processor to control the robot while the other monitors a sensor or takes continuous measurements or to search for an event or value using the robot.

Constant speed: program a path in equal length segments and set a fixed time for each segment.

Interrupts: Real processor interrupts can also be programmed if required without any effect on robot motion. These are all programmed in Roboforth with the same ease as programming the rest of the system. An interrupt can be generated from an input or from an internal timer. For example you might want to take measurements through the analog interface every n milliseconds while the robot program is running and the robot is in motion to build a force/displacement profile for later upload to the PC. You can also embed machine code, that has a dictionary entry and can be included in other definitions.

Heuristics: The robot can be programmed to learn it's own coordinates (self teaching), for example to modify a path you taught it so that it changes as a target position changes. A single path or route can have the same starting position but a variable ending position or vice versa, for example from different tray positions to the same instrument.

Teaching a motion path that can be repeated anywhere in the workspace: If you have the same motion required in a number of different places, for example to pull out sliding shelves from a shelf unit, START-HERE and END-THERE commands allow you to teach the robot one route that can be run from or to different starting or ending positions without having to teach a new path or route for each shelf. Another example would be the robot modifying it's own data according to the results of a sensor search.

Accurate synchronization of motions (for example like playing a piano). Stopwatch function.

Overlays:
Chasing: send coordinates from your PC and have the robot follow them. Send the next coordinates while the robot is still moving. Straight lines with pre-specified maximum deviation from straight. Circles at constant angular speed. Read back position continuously while robot is in motion. Absolute calibration to external datum.

Because Roboforth is based on a real computer language your robot program can include calculations, branching, looping, recursion and so on.

The above is just a glimpse of the immense power of this compact open architecture system, containing several hundred commands and unique extensions. It's a building block approach, which encourages programming interactively with the robot system, a true man-machine interface.

RobWin7

RobWin is a visual project management system for most versions of Windows. (versions available for older computers)

RobWin brings everything together in one screen - the robot controller and your files, the positional information and the procedures that use that information. Once programmed RobWin can be terminated and the new commands sent by other software or equipment or the controller can be set to auto-run with no computer. One window provides the user with direct communication with the robot controller, for entering commands, trying out small robot procedures before including them in the main program.

Features

  • Project Management - maintains all your programming both in the controller and on disk. As locations are learned and edited the data is built up both in the computer and in the controller. RobWin maintains the data files and the header files invisibly and provides a text editing window for the procedure file. Matrix generation is made easy and learned locations can be renamed, edited and so on all on-screen. In addition it can communicate and upload/download data. A typical screen is shown below.
  • Single click commands.
    RobWin generates Roboforth commands, which it sends down to the controller and they appear in the communications window along with your own.
  • Single click learning.
  • Integration of teach pad.
  • Matrix generation (palletizing)
    It's simple - just create the matrix with how many rows and columns you want. Then teach the system 3 corners of the matrix and press "interpolate". RobWin does the rest.
  • Complex path generation.
    Using the curve generator you can piece together paths comprising straight lines and curves of specified angle and radius.

    Typical RobWin7 screen:


    Functions available

    Pull-down menus from left to right:
    File: open ed2 text file, download text file to controller, upload/download binary, print.
    Edit: usual cut and paste etc. for the text files.
    Settings: Load/save a settings file joint names, Cartesian axes, configure system.
    Comm: Change baud rates also change fonts.
    Robot: display or change position of robot in joint or Cartesian coordinates, relative or absolute moves. Choose modes, for example smooth mode, continuous path mode.
    Project: open/save project files.
    Macro: you can write/edit macro commands, which are strings sent to the controller. Ten wide buttons are available for this. In the example above you can see three buttons programmed with the commands ROBOFORTH DE-ENERGIZE and ALIGN.
    View: changes which tool bars show.
    Window: usual Windows feature.

    Buttons, left to right
    Open file, save file, cut, copy, paste, START, CALIBRATE, HOME, TEACH mode - invokes the teach pad, JOG mode - invokes the teach pad in Cartesian mode, GRIP/UNGRIP (toggles), SMOOTH mode, JOINT mode, CARTESIAN mode, upload data from controller, download text to controller.

    A typical open route dialog box

    showing creation of a new matrix TRAY1 in Cartesian mode, 10 columns by 5 rows.

    A typical route editing

    Here a Cartesian matrix, "route" 3SHELF is being edited using add-to-all, which can shift the entire matrix in any direction. This matrix has 10 positions (could just as easily be several hundred) comprising a 4 by 2 matrix plus two extra positions plus a relative position (marked R). The system adds the relative position to the other 10 to provide another 10 displaced from the first (in this case by 30mm in the Z direction), for example as approach or via positions. This economizes on memory space. Only three corners of a matrix need be taught (marked with asterisks) then the interpolate button is pressed and all the in-between positions are computed. RobWin and Roboforth are designed to permit creation and editing of very large matrices with ease and can manage a great number of them to suit the most complex workspace.


    Programming techniques:

    Routes and places may be created using the visual project management system, RobWin. These are put into the dictionary in the controller and the positional data goes into a data area, which holds all the coordinates. The data area is maintained simultaneously in the computer. The system creates a text file for route and place names (header file) and the procedure continues in a second text file. The procedure file is edited in its own window and downloaded to the controller after each edit.
    Normally there are many refinements which are developed to suit any particular application. For example it might be a bad time to invoke the error, which prevents collisions if the robot is inside some machine. So checking for the presence of an object in the way may be done before the robot enters the machine.

    Supervisors

    A supervisor is software that supervises the process running in the controller, perhaps collecting data and usually controls other equipment besides the robot, supervising the entire process. Once RobWin has been used to create spatial data, define procedures, device handling strategies etc. these procedures, known as words, may be sent to the controller via serial link, for example using PRINT commands in BASIC or via RobX ST's ActiveX interface. Proprietary supervisors may be used for example Overlord from Process Analysis Automation (PAA) which has been used successfully to manage and schedule large systems comprising diverse instruments serviced by ST robots. Using RobX you can also write your own supervisor in VB or C. A protocol is given for any other language e.g. Java or Python. Another good approach is to use LabVIEW from National Instruments using RobX.

    RoboForth-GUI

    This is a simple Windows GUI interface to pretty up your final RoboForth program in Windows. It runs under Win32Forth that we can include. It is a single file including communication with the robot and a dialog box that contains within it buttons and a drop down menu. It is easily modified to add extra buttons The whole thing is programmed in Win32Forth and you can easily customize the menu and the names and functions of the buttons in the dialog box. ST will also provide this service for a fixed price. The system can also be extended to a complete Windows application that controls the robot and other instruments in your workspace.

    ActiveX

    If you are writing your own supervisor you can set up your own RS232 handler (or USB/RS232 converter) or can use ST's ActiveX interface. Commands or 'methods' are of the form:
    	short OpenComm(short Port, long BaudRate);
    Opens the communications port. The return value is 1 if successful, 0 for failure. Use the method GetCommErrorString() to get a description of the problem. 
    	short SendString(LPCTSTR String);
    Sends the string to the controller. The string should end with a "Carriage Return" character (ASCII 0D hex). Typical strings will be definitions you have programmed into the controller using Roboforth for example GETPLATE. The return value is 1 if the string was successfully sent, 0 if a communication error occurred. Use the method GetCommErrorString() to get a description of the problem.
    The controller must be in the "ready" state (as reported by GetStatus) before invoking this method. After sending the string invoke GetStatus() until the "ready" state is received. 
    	short GetStatus();
    Gets the controller status as follows:
    -1: Communication error
    0: Waiting (still working on the last command)
    1: Ready, but not OK (some robot error)
    2: Ready, received (and completed) OK

    After receiving status 1 or 2, GetResponse gets the last line sent by the controller. If status is -1, use GetCommErrorString() to get a description of the problem. 

    	CString GetResponse();
    This gets the controller response. If status was 2 the response string indicates what went wrong, for example you told the robot to go to a position it could not reach, or to put an object in a place already occupied.

    Other methods include about box, closing the comms port, getting error codes etc.

    LabVIEW

    Controlling an ST robot from LabVIEW also involves using ActiveX. The ActiveX 'server' is easily accessed from LabVIEW. - see How Can I Access an ActiveX Server in LabVIEW?.

    In addition a whole set of LabVIEW VIs written for us by National Instruments is included with the ST software disk.
    See this example from another ST user.

    Manuals

    Software manuals are provided on disk only; hard copy is no longer used. Manuals are also online at http://strobotics.com/manuals/manuals.htm

    Notes

    * Fourth Generation Languages (4GL) are more popular with AI (artificial intelligence) users than IT (information technology) users who favor third generation languages (3GL). 4GLs tend to be list oriented where as 3GLs are syntax oriented. Examples of 4GLs are: Logo, Lisp (as used by Autodesk ACAD) and Prolog. Examples of 3GLs are: BASIC, COBOL, C, Java.
    For example, compare these identical commands in C (for a different make of robot) and RoboForth:

    C 
    int main(int argc, char **argv){
    Robot k;
    k.calibrate();
    for (int i=0; i<3; i++) {
        k.moveMotAndWait(1, 10000, 100);
        k.moveMotAndWait(1, 11000, 100);
    }
    return 0;
}
    Roboforth 
    : MAIN
CALIBRATE
3 0 DO
  TELL WAIST 10000 MOVETO
  100 MSECS
  TELL WAIST 11000 MOVETO
  100 MSECS
LOOP
;
    C programmers shun Forth, but our users are not programmers. They need something they can use right away and learn as they use it.
    Forth:
    * For more information visit www.forth.com/forth/index.html though ST's Forth was written and compiled by David Sands, and is not a Forth Inc. product.
    * Try Forth yourself
    * Early history of RoboForth

    Online Demonstration