MODULE 2

Module - 2(Sensors, Actuators and Control) Sensor classification- touch, force, proximity, vision
sensors. Internal sensors-Position sensors, velocity sensors, acceleration sensors, Force sensors;
External sensors-contact type, non contact type; Digital Camera - CCD camera - CMOS camera -
Omnidirectional cameras Sensor characteristics. Actuators - DC Motors - H-Bridge - Pulse Width
Modulation - Stepper Motors – Servos, Hydraulic & pneumatic actuators. Control - On-Off Control
- PID Control - Velocity Control and Position Control

2 CLASSIFICATION OF SENSORS
The function of robot sensors may be divided into two principal categories:
1. Internal state.
2. External state.
Internal state sensor:
Internal sensors are used to measure the internal state of a robot. These sensorsdeal with the
detection of variable such as arm joint position, which are used in robot control.
It measures position, velocity, acceleration etc.
External state sensors:
These types of sensors are used to find the robot environment. These sensors deal with the detection of
variables such as range, proximity and touch. External sensing is used for robot guidance, as well as for
object identification and handling. External state sensors may be further classified as:
(i) Contact sensors: These sensors respond to physical contact, such as touch, slip, and
torque.
(ii) Non-contact sensors: These sensors rely on the response of a detector to variations in
acoustic or electromagnetic radiation. The most prominent examples of noncontact sensors measure
range, proximity, and visual properties of an object.

2.1 INTERNAL STATE SENSORS

2.1.a Position and Displacement Sensors

The movement of a joint of a robot may be rotary or linear. Position sensors are used to measure
displacements, both rotary and linear, as well As other robot movements. In many cases, such as in
encoders, the position information may also be used to calculate velocities. The following are common
position sensorsused in robotics:
a) Potentiometers
A potentiometer converts position information into a variable voltage through aresistor. As the sweeper
on the resistor moves due to a change in position, the proportion of the resistance before or after the
point of contact of the sweeper varies.

fig1: potentiometer
Since in this capacity the potentiometer acts as a voltage divider, the outputwill be proportional
to the resistance as

Potentiometers can be rotary or linear and thus can measure linear or rotary motions. Rotary
potentiometers can also be multiple-turn, enabling the user to measure many revolutions of motion.
Potentiometers are either wire wound or thin film deposit (also called conductive plastic), which is a
deposit of a thin film of resistive material on a surface. The major benefit of thin film potentiometers
is that their output is continuous and thus less noisy. As a result, it is possible to electronically
differentiate the output of this type of resistor to find velocity. Potentiometers are generally used as
internal feedback sensors in order to report the position of joints and links. The least count of the
movementthat can be measured is dependent on the tolerance limit of the potentiometer.

a) Encoders
An encoder is a simple device that can output a digital signal for each small portion of a movement.
To do this, the encoder wheel or strip is divided into small sections, as shown in Fig 2. Each section
is either opaque or clear. (It canalso be either reflective or non-reflective.) An LED, on one side
provides a beam of light to the other side of the encoder wheel or strip, where it is seen by a
phototransistor. If the position of the wheel or strip is such that the light can go through the strip, the
phototransistor on the opposite side will be turned on and will have a high signal. If the position of
the wheel or strip is such that the light is blocked, then the phototransistor will get off and its output
will be low (thus, a digital output). As the wheel rotates, it can continuously send signals.
If the signals are counted, the approximate total angular displacement of the wheel can be measured
at any time.
fig.2: Angular and linear encoders

Optical encoders are either opaque wheels with removed material for clear areas (by drilling, cutting,
and otherwise removing) or clear material such as glass with printed opaque areas. Many encoder
wheels are also etched, suchthat they either reflect the light or do not reflect the light. In that case,
the light source and the pick up sensor are both on the same side of the wheel.

There are two basic types of encoders: Incremental and absolute.

Fig.3 are an incremental encoders. In this type of encoder, the areas (arcs) of clear and opaque sections
are all equal and repeating. Since all arcs are the same size, each arc indicates a certain angle of
revolution, which are all the same. If the wheel is divided into only two portions, each portion 180
degrees, its resolution will also be 180 degrees. If the number of divisions increases, the accuracy
increases as well. Thus, the resolution of an optical encoder is related to the number of arcs of clear
and opaque areas. Typical incremental encoders
can have 512 to 1,024 arcs in them, reporting angular displacements with a resolution of 0.7 to 0.35
degree.
An incremental encoder only reports changes to angular position. (It reports thechange in location,
which is the displacement.) However, it cannot report or indicate directly what the actual value of the
position is. In other words, an incremental encoder can only tell how much movement is made, but
unless the initial location is known, the actual position cannot be calculated from the sensor.
Fig 3: output signals of linear encoders Figure 3 shows the output of an incremental encoder. If
only one set of
arcs is used, it will be impossible to determine whether the wheel is rotating clockwise or counter-
clockwise. To remedy this, encoder wheels have two sets of arcs (two channels), 1/2 step out of phase

with each other. As a result, the output signals of the two sets of arcs are also 1/2 step out of phase
with each other. The controller can compare the two signals and determine which one changes from
high to low or vice versa before the other signal. Through this comparison, the controller can be
designed or programmed to determine the direction of rotation. It is actually possible to double the
resolution of the output of incremental encoders without increasing the number of arcs. To do
this, instead of determining the change of signal only at the leading edges from low to high, we can
also count the change at the trailing edges from high tolow, which is equivalent to counting all
changes. With two channels anddouble counting, a resolution four times as good can be achieved.

An alternative to incremental optical encoders is an absolute encoder. Each portion of the encoder
wheel's angular displacement has a unique combination of clear and opaque arcs which give it a
unique signature. Through this unique signature, it is possible to determine the exact position of the
wheel at any time, without the need for a starting position. As shown in Figure 4, there is a multiple
row of arcs, each one different from the others.

fig 4: absolute encoders

Each row must have its own light source and light sensor assembly. Each sensor assembly sends out
one signal. Thus, two rows would require two inputsto the controller (2 bits), three rows would require
3 bits, etc. As can be seen from the figure, an encoder with 4 sets of arcs can have 24 = 16 distinct
combinations, each section covering an angle of 22.5°. To increase the resolution, there would have
to be more arcs, or bits.
a) Linear Variable Differential Transformers (LVDT)
A linear variable differential transformer (or transducer) is actually a transformer whose core moves
along with the distance being measured and thatoutputs a variable analog voltage as a result of this
displacement. Generally ina transformer there are two coils with different number of turns in their
windings as shown in fig 5.

fig 5: LVDT Schematic

The electrical energy into one coil creates a flux, which induces a voltage in the second coil
proportional to the ratio of the number of turns in the windings. The larger the number of turns in the
secondary coil, the larger the voltage is, and, consequently, smaller the current is. However, the
induction of voltage in the secondary is very much a function of the strength of the flux. If no iron
core is present, the flux lines can disperse, reducing the strength of the magnetic field. As a result,
the induction of voltage in the secondary will be minimal. In the presence of an iron core, the flux
lines are gathered inward, increasing the strength of the field and thus the induced voltage. This is
used to
create the output variable voltage in the linear variable differential transformer (LVDT). The output
of an LVDT is very linear, proportional to the input position of the core.

a) Resolvers
Resolvers are very similar to LVDTs in principle, but are used to measure an angular motion instead
of linear motion. A resolver is also a transformer, wherethe primary coil is connected to the rotating
shaft and carries an alternating current through slip rings as shown in fig.6.

fig 6: Resolver schematic Diagram

There are two secondary coils, placed 90° apart from each other. As the rotorrotates, the flux it
develops rotates with it. When the primary coil in the rotor isparallel to either of the two secondary
coils, the voltage induced in that coil is maximum, while the other secondary coil that is
perpendicular to it does notdevelop any voltage. As the rotor rotates, eventually the voltage in
the firstsecondary coil goes to zero, while the second coil develops its maximumvoltage.
For all other angles in between, the two secondary coils develop avoltage proportional to the
sine and cosine of the angle between the primaryand the two secondary coils. Although the output
of a resolver is analog, it is
equal to the sine and cosine of the angle, eliminating the necessity to calculate these values later.
Resolvers are reliable, robust, and accurate.

a) Time-of-Travel (Magneto Reflective) Displacement Sensor

In this sensor, a pulse is sent through a conductor, which bounces back as it readies a magnet. The
time of travel to the magnet and back is converted to the distance if the speed of travel is known. By
attaching the moving part to either the magnet or the conductor, the displacement can be measured.
A simple schematic of the sensor is shown in Fig7.

fig7: Magneto Reflective Displacement Sensor

2.1.b VELOCITY SENSORS

The following are the more common velocity sensors used in robotics. Their application is very much
related to the type of position sensor used. Depending on the type of position sensor used, there may
not even be a need to use a velocity sensor.
i. Encoders
If an encoder is used for displacement measurement, there is in fact no need to use a velocity sensor.
Since encoders send a known number of signals for any given angular displacement, by counting the
number of signals received in a given length of time (dt), we can calculate velocity. A smaller length
of time (dt) yields a more accurate calculated velocity, one that is closer to the true instantaneous
velocity. However, if the rate of rotation of the encoder is low, the velocity measurement may become
inaccurate. This velocity calculation is accomplished by programming the controller to convert
number of signals in a given length of time into velocity information.

ii. Differentiation of Position Signal

If the position signal is clean, it is actually possible, and even simple, todifferentiate the position
signal and to convert it to velocity signal. To do this,it is necessary that the signal be as continuous
as possible in order to prevent creation of large impulses in velocity signal. However, differentiation
of a signal is always noisy and should be done very carefully. Figure below shows asimple R-C
circuit with an opamp that can be used for differentiation, wherethe velocity signal is given by
differentiating the input position signal. Similarly, the velocity (or acceleration) signal can be
integrated to yield position (or velocity) signals.

2.1.a ACCELERATION SENSORS

Accelerometers are very common sensors for measuring accelerations. However, in general,
accelerometers are not used with industrial robots, since no acceleration is generally measured in
these robots. However, recently, acceleration measurements have been used for high-precision
control of linear actuators and for joint feedback control of robots. Acceleration can be calculated by
monitoring speed changes over time, but this approach is rarely used because of accuracy limitations.
A more practical approach is using straingauges to measure the force created by the acceleration of a
known mass. The equation is
∆𝑅𝑆𝐸
𝐹=
𝑅𝐶

where S is surface area, E is Young's modulus, Rand ΔR are the resistance andchange in resistance
of the material, respectively, and C is the deformation constant for the strain gauge material used.
By Newton's law, F = ma, so we can substitute the value of force from equation and solve for a:

The most accurate method for determining acceleration is to measure the

amount of opposing force necessary to maintain a mass in an equilibrium position. Accelerations
ranging from a few tenths of a g to thousands of g's canbe measured this way.
Angular acceleration can be calculated by measuring tangential acceleration ata known distance
from the rotation center and by measuring the rotation angle during a l-s period. The angular
 acceleration is then proportional to the productof the radius and the second derivative of the angle.

2.1.b FORCE AND PRESSURE SENSORS

i. Piezoelectric
Piezoelectric material compresses if exposed to a voltage and produces a voltage if compressed.
This is used in the phonograph to create a voltage fromthe variable pressure caused by the grooves
in the record. Similarly, a piece ofpiezoelectric can be used to measure pressures, or forces, in
robotics. The analog output voltage must be conditioned and amplified for use.
ii. Force-Sensing Resistor
The Force Sensing Resistor (FSR) is a polymer thick-film device that exhibits a decreasing
resistance with increasing force applied normal to its surface.
Force-sensing resistors consist of a conductive polymer, which changes resistance in a predictable
manner following application of force to its surface.
They are normally supplied as a polymer sheet or ink that can be applied
by screen printing. The sensing film consists of both electrically conducting and non-conducting
particles suspended in matrix. The particles are sub- micrometre sizes, and are formulated to reduce
the temperature dependence, improve mechanical properties and increase surface durability.
Applying a force to the surface of the sensing film causes particles to touch the conducting
electrodes, changing the resistance of the film. As with all resistive based sensors, force-sensing
resistors require a relatively simple interface and can operate satisfactorily in moderately hostile
environments. Compared to other force sensors, the advantages of FSRs are their size (thickness
typically less than 0.5 mm), low cost and good shock resistance. A disadvantage is their low
precision: measurement results may differ 10% and more. Force-sensing capacitors offer superior
sensitivity and long term stability, but require more complicated drive electronics.

iii. Strain Gauges

Strain gauges can also be used to measure forces. The output of the strain gauge is a variable
resistance; proportional to the strain, which itself is a function of applied forces. Thus, measuring
the resistance, one can determine
the: applied force. Strain gauges are used to determine the forces at the end effectors and the wrist
of a robot. Figure below is a simple schematic drawing of a strain gauge. Strain gauges are used
within a Wheatstone bridge, as shownin figure.
A balanced Wheatstone bridge would have similar potentials at points A and [Link] the resistance in any
of the four resistors changes, there will be a current flow between the two junctions A and B. Thus,

it is necessary to first calibrate the bridge for zero flow in the galvanometer. Assuming that ft, is the
strain gauge, when under stress, its value will change, causing an imbalance in the Wheatstone bridge
and a current flow between A and B. By carefully adjustingthe resistance of one of the other resistors
until the current flow becomes zero, the change in the resistance of the strain gauge can be determined
from
 2.1 EXTERNAL SENSORS
External sensors are primarily used to learn more about the robot's environment, especially the
object being manipulated. External sensors rangefrom simple limit switches to those measuring
force or temperature.
2.1.a Contact Sensors
To sense objects around it, the robot can use both contact and noncontact sensors. Contact
sensors have the disadvantage that they require physical contact. Thus they cannot give
information about a part of an object that therobot is unable to reach. On the other hand, contact
sensors give important information on object shape, weight, and slippage, which are very useful
in handling objects. These sensors are also generally inexpensive, and the information obtained
requires less processing than visual data.
The simplest contact sensor is an electrical switch. It is easy to use andinexpensive, but it
cannot be used to obtain three-dimensional data or to measure pressure. Pneumatic switches can
also be used, but they are more expensive. By placing contact sensors in arrays, we can obtain
two- dimensional or limited three-dimensional data.
Sensors that must actually touch something in order to operate are limited to circumstances that
will not damage the object or the sensor. One obvious application of an array type of touch sensor
is locating the exact position of ahole so that the robot can insert parts into it.
Tactile sensors are placed on the gripper or other specially designed robot hands. These sensors
can be contact sensing (touch), pressure sensing (force), or slip sensing (slide), and they include
all devices that mimic the operation of
the human hand. In humans, tactile sensation is a property of the skin, althoughthe skin is not
uniformly sensitive. Fingertips, for example, are much more sensitive than the wrists. Touch
sensitivity is also a function of the rate at which the skin is deformed.
The skin responds to separate touches up to a frequency of about 20 Hz, at which point it changes
to a feeling of vibration. Maximum vibration sensitivityis about 250 Hz. Differences between
vibrations frequencies can be determined.
2.1.a Noncontact Sensors
To circumvent possible damage, robots can use proximity sensors, which sensewhen an object is
dose to, but not touching, the robot. Proximity sensors are often used for limit switches. With no
actual mechanical contact required, sensor reliability is increased and safety is improved.
Proximity sensors can bemagnetic or optical. In the magnetic mode, they can sense at a 3/8-5/8-
inches distance by using ferrous actuators. Magnetic techniques include the use of reed relays,
eddy current sensors, and Hall-effect devices. Another type of noncontact sensor is the optical
sensor, which uses the presence of an object toblock out light, and usually operates, in an on/off
mode. More complex opticalsystems can determine the amount of motion through the amount of
light detected. These sensors can be used in conjunction with other techniques to provide force or
pressure information. Optical systems employ visible or infrared (IR) light, and a technique based
upon either light reflection or light interruption is used.
2.2 Robotic drive systems and Actuators
Actuators are the muscles of robots. If you imagine that the links and the joints are the skeleton of the robot,
the actuators act as muscles, which move orrotate the links to change the configuration of robots. The actuator
must have enough power to accelerate and decelerate the links and to carry the loads, yet be light,
economical, accurate, responsive, reliable, and easy to maintain.
s can be: AC motors DC motors Stepper motors Servos DC electric motors are arguably the most
commonly used method for locomotion in mobile robots.
DC MOTORS
DC motors are clean, quiet, and can produce sufficient power for a variety of tasks.
They are much easier to control
The first step when building robot hardware is to select the appropriate motor system. The best choice
is an encapsulated motor combination comprising a:

The first step when building robot hardware is to select the appropriate motor system. The best choice
is an encapsulated motor combination comprising a:

• DC motor
• Gearbox
• Optical or magnetic encoder (dual phase-shifted encoders for detection of speed and direction)
Using encapsulated motor systems has the advantage that the solution is much smaller than that using
separate modules, plus the system is dust-proof and shielded against stray light (required for optical
encoders). The disadvan tage of using a fixed assembly like this is that the gear ratio may only be
changed with difficulty, or not at all. In the worst case, a new motor/gearbox/ encoder combination
has to be used. A magnetic encoder comprises a disk equipped with a number of magnets and one or
two Hall-effect sensors. An optical encoder has a disk with black and white sectors, an LED, and a
reflective or transmissive light sensor. If two sensors are positioned with a phase shift, it is possible
to detect which one is triggered first (using a magnet for magnetic encoders or a bright sector for
optical encoders). This information can be used to determine whether the motor shaft is being turned
clockwise or counterclockwise.
A voltage is applied to the terminals of the motor, which generates a current in the motor armature.
The torque produced by the motor is proportional to the current, and is the motor’s torque constant.
The electrical system of the motor can be modelled by a resistor-inductor pair in series with a voltage
, which corresponds to the back electromotive force. The voltage produced can be approximated as a
linear function of the shaft velocity; is referred to as the back-emf constant
Ve = Keω

Servos
Servos are not servo motors! DC motors are sometimes also referred to as “servo motors”. This is not
what we mean by the term “servo”. A servo motor is a high-quality DC motor that qualifies to be
used in a “servoing application”, i.e. in a closed control loop. Such a motor must be able to handle
fast changes in position, speed, and accel eration, and must be rated for high intermittent torque
 A servo, on the contrary, is a DC motor with encapsulated electronics for PW control.
A servo has three wires: VCC, ground, and the PW input control signal. Unlike PWM for DC motors,
the input pulse signal for servos is not trans formed into a velocity. Instead, it is an analog control
input to specify the desired position of the servo’s rotating disk head. A servo’s disk cannot per form
a continuous rotation like a DC motor. It only has a range of about ±120° from its middle position.
they do not provide any feedback to the outside.

CONTROL
On-Off Control
In the simplest case, power to the motor is either switched on (when the speed is too low) or switched off
(when the speed is too high). This control law is represented by the formula below, with:
R(t) motor output function over time t
vact(t) actual measured motor speed at time t
vdes(t) desired motor speed at time t
KC constant control value

What has been defined here, is the concept of an on-off controller, also known as “piecewise constant
controller” or “bang-bang controller”. The motor input is set to constant value KC if the measured velocity is
too low, otherwise it is set to zero. Note that this controller only works for a positive value of vdes.

PID Control
PID = P + I + D The simplest method of control is not always the best. A more advanced con troller and
almost industry standard is the PID controller. It comprises a proportional, an integral, and a derivative control
part.

Proportional Controller
For many control applications, the abrupt change between a fixed motor control value and zero does not result
in a smooth control behavior. We can improve this by using a linear or proportional term instead. The formula
for the proportional controller (P controller) is:

R(t) = KP · (vdes(t) – vact(t))

The difference between the desired and actual speed is called the “error function”.
Varying the “controller gain” KP will change the controller behavior. The higher the KP chosen, the faster the
controller responds; however, a too high value will lead to an undesirable oscillat ing system. Therefore it is
important to choose a value for KP that guarantees a fast response but does not lead the control system to
overshoot too much or even oscillate.
 Each P controller will keep a certain “steady-state error” from the desired velocity, depending on the
controller gain KP

Integral Controller
Unlike the P controller, the I controller (integral controller) is rarely used alone, but mostly in combination
with the P or PD controller.
The idea for the I controller is to reduce the steady-state error of the P controller.
With an additional integral term, this steady-state error can be reduced to zero, as seen in the measurements
The equilibrium state is reached somewhat later than with a pure P controller, but the steady-state error has
been eliminated.
When using e(t) as the error function, the formula for the PI controller is:

R(t) = KP · [ e(t) + 1/TI · 0∫ t e(t)dt ]

We rewrite this formula by substituting QI = KP/TI, so we receive independent additive terms for P and I:

R(t) = KP · e(t) + QI · 0∫ t e(t)dt

Derivative Controller

Similar to the I controller, the D controller (derivative controller) is rarely used by itself, but mostly in
combination with the P or PI controller.
The idea for adding a derivative term is to speed up the P controller’s response to a change of input.
The PD controller reaches equilibrium faster than the P controller, but still has a steady-state error. The full
PID controller combines the advantages of PI and PD. It has a fast response and suffers no steady-state error.
When using e(t) as the error function, the formula for a combined PD controller is:

R(t) = KP · [ e(t) + TD · de(t)/dt]

The formula for the full PID controller is:

R(t) = KP · [ e(t) + 1/TI · 0 ∫ t e(t)dt + TD · de(t)/dt ]