Introduction to Robotics

Narendiran Anandan
Serkan Ergun
Robot Paradigms
• A paradigm is a philosophy or set of assumptions and/or techniques
which characterise an approach to a class of problems.

• There are currently three paradigms for organising intelligence in

robots.
• Hierarchical
• Reactive
• Hybrid

• They are generally described by the relationship between the three

commonly accepted primitives of robotics. SENSE, PLAN, ACT.
SENSE, PLAN, ACT
ROBOT PRIMITIVE INPUT OUTPUT
SENSE Sensor data Sensed information
PLAN Information (Sensed, cognitive) Directives
ACT Directives Actuator commands

SENSE:
Takes the raw readings from the sensor and converts it into meaningful
information about the world.

PLAN:
Based on sensed information and goals, generate directives.

ACT:
Based on directives generate actuator commands.
Hierarchical Paradigm

• Oldest Paradigm.
• Prevalent from 1967-1990.
• This was based on an introspective view of how people think.
• Robot senses the world, plans next action and acts.
• At each loop, the robot explicitly plans the next action.
• All sensing data tends to be gathered into one global world model.
• Constructing generic global models is a difficult problem.
Reactive Paradigm

• In this paradigm, the sensed information is directly coupled to

output actions.
• Started as alternative approach to Hierarchical paradigm.
• Heavily used in robotics research from 1988 to 1992.
• Used by AI researchers to investigate biology, and intelligence.
• Helped by rapidly decreasing cost of computer hardware and
increase in computing power.
• Researchers could emulate frog and insect behaviour for about 500$.
Hybrid Paradigm

• Emerged in the 1990s.

• Current area of research.
• The robot first plans how to best decompose
the task into a set of subtasks (mission
planning), and then determines what are the
suitable behaviour for accomplishing each
subtask.
• Then the behaviours start executing in
parallel.
Sensors and Actuators

• Sensors are device that produce an • Light

output signal in response to some • Temperature
physical parameter in the
environment. • Distance
• Velocity
• An actuator is a device that • Acceleration
produces a motion by converting • Angle
energy and signals going into the • Sound
system.
• More …
Sensor

Environmen Interface Output

tal Variable Signal
with Sensing Interface Output
Conditioner
Environmen Element (usually Signal
t (Optional)
(Measurand) electrical)
Sensor

Environmen Interface Output

tal Variable Signal
with Sensing Interface Output
Conditioner
Environmen Element (usually Signal
t (Optional)
(Measurand) electrical)
Sensor Classification
• They can be classified in a number of ways.
• Active or Passive
• Contact or Noncontact
• Absolute or relative
• Function
• Physical phenomena
Sensor Attributes - Measurand
• Positional and Dimensional – linear displacement, angular displacement,
liquid level, area, volume
• Proximity – Metal detection
• Static Mechanical – Mass, Pressure, Torque
• Dynamic Mechanical – Velocity, acceleration, flow rate, acceleration
• Thermal – Temperature, Temperature Variation
• Electromagnetic – Wavelength, Frequency, Intensity, Polarisation
(particularly optical and infrared)
• Electrostatic – Charge, Capacitance, Resistance
• Magnetic – Flux
• Sonic/Acoustic – Sound frequency, Intensity and direction
Sensor Attributes – Operating Distance

• Remote Sensing – Few meters to millions of kilometres.

• Only option when the measurand is always distant. (Astronomical and
navigational)
• Background noise is present when large distance are involved.

• Near contact Sensing – Few millimetres to few meters.

• Contact Sensing – Touching the measurand.

Sensor Attributes – Invasiveness
• Sensor Invasiveness: Degree to which the sensor reacts with the
measurand environment in which it is placed.

• Measurand Invasiveness: Measurand can damage the sensor in a

hostile environment. (Radiation, acidity, high temperature, chemicals)
Sensor Attributes – Form of Output Signal
• Mechanical Decimal Binary Gray
0 0000 0000
• Electrical 1 0001 0001
• Analog 2 0010 0011
• Voltage or Current dependent 3 0011 0010
• Modulated by frequency or 4 0100 0110
amplitude
5 0101 0111
• Digital
6 0110 0101
• Binary coded
7 0111 0100
• Gray coded
8 1000 1100
9 1001 1101
10 1010 1111
11 1011 1110
12 1100 1010
13 1101 1011
14 1110 1001
Sensor Performance Characteristics
• Linearity/Nonlinearity
Output

Output
Output
Input Input Output

𝑦 =1.2 𝑥 𝑦 =0.1 𝑥 2 +0.03 𝑥 𝑦 =1.5 𝑥 2 − 40.7 𝑥 +11006

Sensor Performance Characteristics
• Range of operational accuracy

• Repeatability and Reproducibility

• Drift, Hysteresis

Output
• Responsiveness

• Reliability
Input
Sensor ergonomic and economic factors
• Human machine interfaces
• Ease of installation and repairing
• Cost
• Packaging
• Market Volume
Displacement Sensing – Resistive

𝑙
𝑅=𝜌
𝐴
Displacement Sensing – Resistive

• Advantages:
• Easy to use
• Low cost
• Simple circuit
• High amplitude output
signal
• Proven technology
• Disadvantages
• Limited bandwidth
• Frictional loading
• Inertial loading
• Wear and Tear
Displacement Sensor – Inductive
Single Coil Variable Reluctance Displacement Sensor
A basic inductive sensor consists of a
magnetic circuit made from a ferromagnetic
core with a coil
wound on it.

The coil generates and drives flux through

the magnetic circuit through the core and the
air gap.

The presence of the air gap causes a large

increase in circuit reluctance and a
𝐿0 corresponding decrease in the flux.
𝐿=
1+𝛼 𝑑 Hence, small variations in the air gap result in
Displacement Sensor – Inductive
Differential Variable Reluctance Displacement Sensor
A linear output can be obtained from L1
and L2 for small displacements using
appropriate circuits.

Continuous resolution

High output

Affected by external magnetic fields

𝐿10 A typical sensor of this type has an input

𝐿1 =
1+ 𝛼 ( 𝑑 + 𝑥 ) span of 1 cm, a coil inductance of 25 mH,
𝐿2 =
𝐿20 and a coil resistance of 75 Ω.
1+ 𝛼 ( 𝑑 − 𝑥 )
 Linear Variable Differential Transformer (LVDT)

±0.25 mm to ± 7.5 cm range.

-265°C to 600°C operating temperature.
Radiation resistant designs are available.
Displacement Sensor – Inductive
Eddy Current Sensors
Excitation in the sensor coil induces eddy current in the target
surface.

The eddy currents generate opposing magnetic field that affects

coil inductance.

Change in inductance can be measured to obtain target distance.

Range: 0.4 to 80mm

Resolution: 0.5 μm
High linearity and temperature stability.
High speed applications.

Since the target material, shape, etc , influence the output, it is

necessary to calibrate the system statistically for a specific target.
Proximity Sensing – Capacitive
• Capacitance is the ability of a
component or circuit to collect and
store energy in the form of an
electrical charge.
𝐴
𝐶=𝜀
𝑑
Displacement Sensor – Capacitive

𝐴
𝐶=𝜀
𝑥

𝐴 − 𝑤𝑥
𝐶=𝜀
𝑑
𝑤

𝐴
Displacement Sensor – Capacitive
Commercial Capacitive displacement
sensors from Micro-Epsilon

Non contact

High Accuracy

Vacuum and clean room applications.

Immune to magnetic field interference.

Ultrasound Displacement Transducer

Transducer Types
Mechanical – Whistel, Sirens
Electromagnetic – Microphone
Piezoelectric – Quartz, Ceramics, etc
Electrostatic – Capacitor with free
membrane

Time of Flight Techniques

Pulse Echo, Phase angle, Frequency
Optical Encoder Displacement Sensors
• Optical encoders are used to measure either
angular or linear positions.

• Those used for angular detection are commonly

called rotary or shaft encoders.

• They use light as the means to transform

movement into electrical signals.

• All devices have two basic building blocks: a main

grating and a detection system. It is the position
of one with respect to the other that is detected.

• Types: Incremental or Absolute.

Optical Encoder Displacement Sensors – Absolute
Encoders
Direct Binary Encoder
Has N tracks, each providing one
bit of the a direct binary number.

For each track the readout unit

consist of a light source, a mask
and a photodetector.

As the mask moves over a clear

region on the grating, the
photodetector output increases
and decreases.

Truncated triangular wave output

is converted to square wave output
Optical Encoder Displacement Sensors – Absolute
Encoders
7 0111
8 1000

7 ? 8 7 9 8 7 13 8

100 110
Optical Encoder Displacement Sensors – Absolute
Encoders

Decima Decima
Binary Gray Binary Gray
l l
0 0000 0000 8 1000 1100
1 0001 0001 9 1001 1101
2 0010 0011 10 1010 1111
3 0011 0010 11 1011 1110
4 0100 0110 12 1100 1010
5 0101 0111 13 1101 1011
6 0110 0101 14 1110 1001

Vee Scan Gray Code Encoder

7 0111 0100 15 1111 1000
MEMS Accelerometers and Gyroscope

MEMS: Micro-
ElectroMechanical
System
MEMS Accelerometer

Accelerometer Grade Main Application Bandwidth g-Range

Consumer Motion, static acceleration 0 Hz 1g

Automotive Crash/stability 100 Hz <200 g/2 g

Industrial Platform stability/tilt 5 Hz to 500 Hz 25 g

Tactical Weapons/craft navigation <1 kHz 8g

Submarine/craft
Navigation >300 Hz 15 g
navigation
FMCW Radar
• Can measure very small ranges to the
target (the minimal measured range is
Chirp
comparable to the transmitted
wavelength).
• Can measure simultaneously the target
range and its relative velocity.
• Very high accuracy of range measurement.
• Signal processing after mixing is
performed at a low frequency range,
considerably simplifying the realization of
the processing circuits.
• Safety from the absence of the pulse
radiation with a high peak power.
FMCW Radar

2𝑑
τ=
𝑥1 𝑐
𝑥1
𝐴 sin ( 2 𝜋 𝑓 0 𝑡 + 𝜃 0 )
𝑥2 𝑥 𝑜𝑢𝑡
2 𝑆𝑑
𝑓 0=
𝑐

𝑥 1 =sin ( 𝜔 1 𝑡 + 𝜃 1 ) 4𝜋𝑑
𝜃0=
𝑥 2 =sin ( 𝜔 2 𝑡 + 𝜃 2 ) 𝜆

𝑥 𝑜𝑢𝑡 =sin [ ( 𝜔 1 − 𝜔 2 ) 𝑡 +( 𝜃 1 − 𝜃 2 ) ]
Global Navigation: Global Positioning System (GPS)

GPS is a system of 30+ navigation

satellites circling Earth.

We know where they are because

they constantly send out signals.

A GPS receiver in listens for these

signals.

Once the receiver calculates its

distance from four or more GPS
satellites, it can figure out where
you are using trilateration.
Global Navigation: GLONASS, Galileo, BeiDou
• GPS emerged from the US military, and made publicly available.
• Accuracy for consumers: few meters, For military: (sub-)millimeter

• GLONASS : Russian
• GALILEO: EU
• BeiDou: China
• IRNSS: India (in planning), QZSS: Japan (in planning)

• Most devices use a combination of the navigations systems

• Local Restrictions (Devices in Russia must have GLONASS enabled, devices in
China need BeiDou)
Light Sensor (reverse LED)

Photoresistors

Phototransistor

Photodiodes
Laser Range Finder
• Measurement Method: Time of
flight.
• Range: few meters to few
kilometers
• Resolution 10mm
• Line of sight measurement
• Expensive
• Application:
• Production and Control
• 3D mapping
• High precision guidance
Vision
• Vision sensors provide the
richest information
• Single camera / Stereo Camera
• Geometric information
• Texture
• Colour
• Applications
• Distance measurement
• Object detection recognition
• Control