Classification of Robotics Languages:

▪ First Generation Language

▪ Second Generation Language
▪ Future Generation language
 First Generation Language
▪ The first generation language uses a
combination of command statement and teach
pendent procedures for developing robot
programs.
▪ The features of this language include the
ability to define manipulator motions, straight
line interpolation and sensor commands.
▪ VAL is an example of first generation
language,
 Programming Procedure

Edit Save Test Optimize Store

Program Program Program Program Program
Programming lanuages

Programming languages:

ABB robot language = RAPID

STÄUBLI robot language = Val 3

KUKA robot language = KRL

 Programming KUKA Robots with KRL

Source: KUKA
 Robot Programming: KUKA Teach Box
Motion mode

Emergency Stop

Automatic mode: STOP 6D mouse

Soft key

Functional window Graphical display

Jogging
Axis movement
KUKA Teach Box ( rear view)
Safety switch
KRL Program language: Basic Instructions

-Motion
PTP HOME
PTP point_x,
LIN point_x, (CP-linear)
CIRC point _x, (CP-circular),
-Motion specific parameters
VEL <Wert> speed [m/s]
VEL <%> with PTP
CONT pass over
Basic instructions -Motion data
(KRL= KUKA Robot Language) PDAT PTP motion data
CPDAT CP motion data
-Program identifier: DEF < Name>
-Program-Initialisation: INI
-Program-End: END
KRL Programming language

Sample of KRL-Program (KUKA):

1 DEF my program
2 INI
3 PTP HOME VEL=100%,
4 PTP P1 VEL= 60% PDAT1 TOOL[2]
5 LIN P2 VEL= 1m/s CPDAT1 TOOL[2]
6 PTP P3 VEL= 80% PDAT1 TOOL[2]
7CIRC P4,P5 VEL=1,5m/s CPDAT1 TOOL[2]
8 .....
9 .......
10 PTP HOME VEL=100%
11 END
Second Generation Language

▪ The second generation language overcome

many of the limitations of the first generation
language.
▪ This enables the robot to accomplish more
complex tasks.
▪ VAL II, RAIL and AML are the examples of
second generation languages.
 Robot-oriented programming
 “second generation” languages: structured programming
with characteristics of an interpreted language (interactive
programming environment)
 typical instructions of high-level languages are present
(e.g., logical branching and while loops)
 ad-hoc structured robot programming languages (more common)
 development of robotic libraries in standard languages (preferred)
 access to the action level
 handle more complex applications where the robot needs to
cooperate/synchronize with other machines in a work cell
 examples of languages: VAL II (Unimation), AML (IBM),
PDL 2 (Comau), KRL (KUKA)

Robotics 13
1
 KUKA user interfaces

Teach pendant
KRL
programming
Ethernet RSI XML

 Fast Research
Interface

Robotics 14
1
 KRL language
 basic instruction
set:

DO

 basic data set: frames, vectors +

DECLaration
Robotics 15
1
 KRL language
 typical motion
primitives

PTP motion LIN motion CIRC motion

(point-to-point, linear (linear in (circular in
in joint space) Cartesian space) Cartesian space)

end-effector
orientation
PTP motion
CONST orientation
(linear in RPY angles)

Robotics 16
1
 KRL language
 multiple coordinate frames (in Cartesian space) and
jogging of robot joints

Robotics 17
1
 KUKA Ethernet RSI
Robot Sensor Interface

cyclical data transmission from the robot controller to an external

system (e.g., position data, axis angles, operating mode, etc.) and
vice versa (e.g., sensor data) in the interpolation cycle of 12 ms
influencing the robot in the interpolation cycle by means of an
external program
direct intervention in the path planning of the robot
recording/diagnosis of internal signals
communication module with access to
standard Ethernet via TCP/IP protocol as
XML strings (real-time capable link)
freely definable inputs and outputs of the
communication object
data exchange timeout monitoring

Robotics 18
1
 Example of RSI use -1

 deburring task with robot motion controlled by a force sensor

䙵 work piece to be deburred along the
edge under force control
䙶 tool with force sensor
© robot
䙸 robot controller
FX measured force in the X direction of
the BASE coordinate system
(perpendicular to the programmed path)
v direction of motion

ZBASE

LIN_REL = linear Cartesian path relative

YBASE to an initial position (specified here by the
force sensor signal)
XBASE

Robotics 19
1
 Fast Research Interface (FRI)
for KUKA Light Weight Robot (LWR-IV)

UDP socket communication up to 1 KHz (1÷100 ms cycle time)

here, we develop
our C++/ROS code for:
• trajectory planning
• kinematic control
• redundancy resolution
• torque/dynamic control
• physical HRI
• ...

available
at DIAG
Robotics Lab
since Sep 2012

Robotics 20
1
▪ The features and capabilities of second
generation languages are as follows:
1. Motion Control
2. Advanced sensor capabilities
3. Limited intelligence
4. Communication and data processing
Future Generation Language

▪ Future generation language involves a concept

of world modelling.
▪ The robot possesses knowledge of three
dimensional world and is capable of developing
its own step by step procedure to perform a task
based on a stated objective of what is to be
accomplished.
 Task-oriented programming

 “third generation” languages (for research, not yet

available on the market)
 similar to object-oriented programming
 task specified by high-level instructions performing actions
on the parts present in the scene (artificial intelligence)
 understanding and reasoning about a dynamic environment
around the robot
 access to the task level

Robotics 23
1
 Functional control architecture
reference model
sensor knowledge decision
processing models strategies
task
S M D
level
. . . operator

. action
global memory

S M D

interface
level

. primitives
S M D
level
. . .
. servo
S M D
level
sensors actuators
Robotics 24
1
 Functional architecture: Modules
horizontal decomposition reference model
sensor knowledge decision
processing models strategies
task
S M D
level
. . . operator

. SENSORY MODULES
action i
global memory

S M D

nt er c e
acquisition, processing
leveland
fa
integration of sensory data
. primitives
S M D
level
. . .
. servo
S M D
level
sensors actuators
Robotics 25
1
 Functional architecture: Modules
horizontal decomposition reference model
sensor knowledge MODELING MODULES
decision
processing models strategies
a priori knowledge about
robot task
S M D + environment
level system,
. . updated using information
. operator
from sensory modules
. action
global memory

S M D

interface
level

. primitives
S M D
level
. . .
. servo
S M D
level
sensors actuators
Robotics 26
1
 Functional architecture: Modules
horizontal decomposition reference model
sensor knowledge decision
processing models strategies
DECISION MODULES task
M D
• decompositionS(in time and space) level
.S .M level . operator
oflgtasks. into actions of lower
ob
• choice and processing of strategies D action

interface
level
al memory

. primitives
S M D
level
. . .
. servo
S M D
level
sensors actuators
Robotics 27
1
 Functional architecture: Modules
horizontal decomposition reference model
sensor knowledge decision
processing models strategies
task
S M D level
. GLOBAL. MEMORY . operator
data and information relevant
. action
global memory

S M D
to all levels (updated estimate leel

interface
of robot + environment state) v
. primitives
S M D
level
. . .
. servo
S M D
level
sensors actuators
Robotics 28
1
 Functional architecture: Modules
horizontal decomposition reference model
sensor knowledge decision
processing models strategies
task
S M D level
. OPERATOR
. INTERFACE
. operator
allows intervention by an action
.
global memory

S
operator atM any level of
D
the leel

interface
functional hierarchy v
. primitives
S M D
level
. . .
. servo
S M D
level
sensors actuators
Robotics 29
1
 Reference model: Levels

 task level: objective of the task (as specified by the

TEMPORAL CONSTRAINTS
user) analyzed and decomposed into actions (based
INFORMATION COMPLEXITY

on knowledge models about the robot and the

environment systems)
 action level: symbolic commands converted into
sequences of intermediate configurations
 primitives level: reference trajectories generation for
the servo level, choice of a control strategy
 servo level: implementation of control algorithms,
real-time computation of driving commands for the
actuating servomotors

Robotics 30
1
 A functional
architecture for
industrial robots D action
reference frames
request state path points
interpolation modes
camera
S M D
primitives
. ..
qdes qdes qdes
data control algorithm
request state

S M D
servo

data actuator
commands
force velocity position
Robotics 1
31
 A functional
architecture for
industrial robots D action
vertical decomposition
reference frames
request
ACTION LEVEL state path points
interpolation modes
camera
• interpreter
S of high-level M
commands D
primitives
• task decomposition made by human . ..
qdes qdes qdes
operator data control algorithm
• no sensory and modeling modules
request state
(unless a multi-modal cognitive
S M D
human-robot interaction is possible) servo

data actuator
commands
force velocity position
Robotics 1
32
 A functional
architecture for
industrial robots
D action
vertical decomposition
reference frames
request state path points
interpolation modes
camera
S M D
primitives
. ..
qdes qdes qdes
data control algorithm
PRIMITIVES LEVEL
request state
• S: (only for an active interaction with the environment)
world geometry, Sinteraction stateM D
servo
• M: direct and inverse kinematics, dynamic models
• D: command encoding, path generation, trajectory
interpolation, kinematic
data inversion, analysis of servoactuator
state, emergency handling commands
force velocity position

Robotics 33
1
 A functional
architecture for
industrial robots D action
vertical decomposition
reference
SERVO LEVEL frames
request state path points
• S: signal conditioning, internal state of manipulator, state of
interpolation
camera
interaction with environment
modes

• M: direct kinematics,
S Jacobian, inverse
M dynamics
D
primitives
• D: command encoding, micro-interpolation, error handling, .
digital control laws, servo interface .
.
q q
S M D des des
servo
qdes
data control
algorithm
data actuator
request state commands
force velocity position

Robotics 34
1
 Interaction among modules

task execution
motor control
perception horizontal
modeling
planning
sensors actuators activation
(sequential)

plan changes to world

identify object
monitor changes vertical
sensors build map actuators activation on demand
(subsumption)
explore
wander
avoid obstacles

Robotics 35
1
 LAAS architecture

 alternative example by
LAAS/CNRS in Toulouse
 five levels
 decision
 execution (synchronization)
Functional

functional (modules)
Level

 logical for interface

 physical devices

R. Alami et al.
“An Architecture for Autonomy,”
Int. J. of Robotics Research, 1998

Robotics 36
1
ROBOT LANGUAGE STRUCTURE
▪ Powered leadthrough and manual leadthrough are
two methods of programming.
▪ The difference between the two is in the manner
in which the manipulator is moved through the
motion cycle during programming.
▪ Powered leadthrough is commonly used as thc
programming method for playback robots with
point-to-point control.
▪ It involves the use of a teach pendant (hand-held
control box) that has toggle switches and/or
contact buttons for controlling the movement of
the manipulator joints.
▪ Manual leadthrough programming is
convenient for programming playback robots
with continuous path control where the
continuous path is an irregular motion pattern
such as in spray painting.
▪ This programming method requires the
operator to physically grasp the end-of. arm or
tool attached to the arm and manually move it
through the motion sequence, recording the
path into memory.