Chapter 6 : Automated and emerging

technologies
Prepared by Mr. Ala abu jaber
Objectives :
6.1 Automated systems
6.1 Automated systems
6.1.1 Sensors, microprocessors and actuators
An automated system is a combination of software and hardware (for
example,
sensors, microprocessors and actuators) that is designed and
programmed to
work automatically without the need of any human intervention.
However, such
systems often involve human monitoring.
6.1.1 Sensors, microprocessors and actuators (Cont..)
The role of sensors, microprocessors and actuators was discussed at great
length in Section 3.2. It may be worth the reader revisiting this part of the
book before continuing with this chapter; you should remember that:

» Sensors are input devices that take readings from their surroundings and
send this data to a microprocessor or computer. If the data is analogue, it is
first converted into a digital format by an analogue-digital converter (ADC).
6.1.1 Sensors, microprocessors and actuators (Cont..)

» The microprocessor will process the data and take the necessary
action based on programming.
» This will involve some form of output, usually involving signals sent
to actuators to control motors, wheels, solenoids, and so on.
Advice
» On first sight, all of the examples in 6.1 will appear very complicated.

» However, you will not learn any of the industrial or scientific processes described
fully in this chapter. Any processes used in any questions will be fully described to
you (possibly including a diagram).

» You basically need to go through each example carefully and understand the

processes taking place. In other words, what is the interaction between the
sensors, actuators and microprocessor/computer to allow the process to take
place.
Advice
» On completion of Section 6.1 the important learning process is to understand

the sensor, actuator and microprocessor interaction; to this end, you need to do

two things:

– for each example, complete a table as follows:

6.1.2 Advantages and disadvantages of automated systems

We will show the role of sensors, microprocessors (or computers)

and actuators in the following application areas:
1. » industrial
2. » transport
3. » agriculture
4. » weather
5. » gaming
6. » lighting
7. » science.
1.Industrial applications
Automated systems are used in a number of industrial applications. Many of
the automated systems involve robotics, which is covered in more depth in
Section 6.2.
In recent years, the focus on increased automation has led to improved quality
and flexibility. For example, in the manufacture of car engines, when done
manually, the installation of pistons into the engine had an error-rate of ~1.5%;
with automated systems, the error-rate has fallen to 0.00001%.
We will now consider two very different industrial applications:
A key use of automated systems is in the control and monitoring of a nuclear
power station. This is a good example, since automation gives increased
safety both in the process itself and to the workforce. At the centre of the
system is a distributed control system (DCS). DCS is essentially a powerful
computer that has been programmed to monitor and control the whole
process with no human interaction required:
Example 1: A nuclear power station (Cont..)

Data from a number of sensors is sent to a DCS (computer) – if the data is

analogue, it must first be converted into digital format using an ADC. The
DCS will have access to a large database containing operational data and
parameters. If any action needs to be taken, then signals will be sent to the
appropriate actuators to operate pumps, valves or even an emergency
shutdown system. The key here is that the system is fully automated. A
human operator (the supervisor) will sit in a remote control room where a
schematic of the process will show on a large screen. While the process is
fully automatic, the supervisor can still override the DCS and shut down the
process.
The main advantages of this automated system are:

1. » much faster than a human operator to take any necessary action

2. » much safer (an automated system is more likely to make timely
interventions than a human; it also keeps humans away from a
dangerous environment)
3. » the process is more likely to run under optimum conditions since
any small changes needed can be identified very quickly and action
taken.
4. » in the long run, it is less expensive (an automatic system replaces
most of the workforce who would need to monitor the process 24
hours a day).
The main disadvantages of this automated system are:

1. » expensive to set up in the first place and needs considerable testing

2. » always possible for a set of conditions to occur that were never
considered during testing which could have safety implications (hence the
need for a supervisor)
3. » any computerised system is subject to cyberattacks no matter how
good the system (one way round this is to have no external links to the
DCS; although the weak link could potentially be the connection to the
supervisor)
4. » automated systems always need enhanced maintenance which can be
expensive.
This automated system also depends on sensors, a computer, actuators
and software. Process 1 is the manufacture of the paracetamol. Process 2
is the making of the solid tablets. Both processes are monitored by a
number of sensors that send their data back to a central computer. The
computer consults its database to ensure both processes are operating within
correct parameters. Any necessary action is taken by the computer, sending
signals to the appropriate actuator to operate pumps,valves, heaters,
stirrers or pistons to ensure both processes can operate without any human
intervention. Again, this system uses a remote monitoring station manned by
an operator. The system is fully automated, but the operator can override the
central computer system if necessary.
The main advantages of this automated system are:

1. » much faster than a human operator to take any necessary action

2. » much safer (an automated system is more likely to make timely
interventions than a human if necessary; it also keeps humans away from a
potentially dangerous environment)
3. » the process is more likely to run under optimum conditions since any
small changes needed can be identified very quickly and action taken
4. » in the long run, it is less expensive (an automatic system replaces most
of the workforce who would need to monitor the process 24 hours a day)
5. » more efficient use of materials
6. » higher productivity
7. » more consistent results.
The main disadvantages of this automated system are:

1. » expensive to set up in the first place and needs considerable

testing
2. » always possible for a set of conditions to occur that were never
considered during testing which could have safety implications (hence
the need for a monitoring station)
3. » automated systems always need enhanced maintenance which
can be expensive
4. » any computerised system is subject to cyberattacks no matter how
good the system.
2.Transport
As with industrial processes, many of the automated systems in transport refer

to robotic systems (for example, autonomous buses/cars, autonomous trains and

unpiloted aircraft). These will be considered in Section 6.2.

But automated systems are still used in manually controlled transport, which

includes cars, buses/lorries, trains and aircraft. (Examples 3 and 4 which follow,

will use cars as the application.)

2.Transport (Cont..)

For example, modern trains will use an automatic signal control

system. If the driver of the train goes through a red (stop) light, then the
computer will automatically stop the train. This will make use of sensors
at the side of the track sending signals to the on-board computer;
actuators will be used to apply the brakes. Airplanes extensively use
automatic pilots, which control the wing flaps, throttle and rudder to
maintain the correct height, speed and direction.
Example 3: Self-parking cars
The driver goes along the row of parked cars. On-board sensors and cameras
gauge the size of any parking spaces, and the on-board computer warns the
driver if a suitable space has been found. The driver then selects auto-parking
and the on-board computer takes over. Actuators are used to operate the
steering rack, brakes and throttle under the full control of the computer. This
allows the car in Figure 6.4 to go from Step 1 to Step 2 automatically and
complete the parking manoeuvre with no driver intervention.
Example 3: Self-parking cars
Sensors in the bumpers of the car are both transmitters and receivers. The
sensors transmit signals that bounce off objects and are reflected back. The
car’s on-board computer uses the amount of time it takes for the signal to return
to the sensor to calculate the position of any objects. The sensors give the
computer a 3D image of its surroundings. This allows the car to fit into its
parking space automatically with no driver intervention. (Note: cheaper and older
self-parking systems are not fully automatic; they require the driver to operate
the brakes and throttle manually and only control the steering.)
The main advantages of this automated system are:

1. » allows the same number of cars to use fewer parking spaces

2. » avoids traffic disruption in cities (a manually controlled car takes
several seconds to fit into a parking space)
3. » cars can fit into smaller spaces
4. » fewer dents and scratches to cars (reduced insurance claims)
5. » safer system since sensors monitor all objects, including young
children (the car’s manoeuvre will be stopped if any new object is
encountered)
6. » very consistent results.
The main disadvantages of this automated system are:

1. » over-reliance on automated systems by the driver (loss of skills)

2. » faulty/dirty sensors or cameras can send false data/images to the
on-board computer which could lead to a malfunction
3. » kerbing of wheels is a common problem since the sensors may not
pick-up low kerbs
4. » expensive option that doesn’t really save the driver any money
5. » requires additional maintenance to ensure it functions correctly at
all times.
Adaptive cruise control

Adaptive cruise control makes use of sensors, an on-board computer

and actuators to allow a car to remain a safe distance from another
vehicle.
The driver will set a cruising speed (for example, 100 kph) on his touch
screen in the car. Lasers (set into the bumpers of the car) are used to
send out signals constantly.
Adaptive cruise control (Cont..)

The lasers bounce off the vehicle in front of the car and are reflected
back to the car’s sensors. The time taken for the signal to bounce back is
used by the on-board computer to calculate the distance between the
two vehicles. If the car is getting too close to the vehicle in front, the
computer will send signals to slow the car down.
Adaptive cruise control (Cont..)

This is done by actuators applying the brakes and/or reducing the throttle. If the
distance between vehicles is greater than the safe distance, the computer will

check to see if the current speed equals the value set by the driver. If the speed

is different to the set speed, the computer sends signals to the actuators to

increase or decrease the throttle.

3.Agriculture

There are many examples of the use of automated systems in agriculture.

Again, many of the systems involve robotics, which is fully described in
Section 6.2. We will now consider one important example that is being
used in Brazil to irrigate crops automatically.
The watering of crops (irrigation) is fully automatic and also involves
considerable amounts of wireless transmission. This allows the system to
be used in very remote areas that are vast in size – some of the fields are
more than 10 km2 in area.
Data from an automatic weather station (see next example) is received by the
controller (a computer system) every ten minutes. This is particularly important if
very wet or very dry conditions are being predicted or detected by the weather
station. Ultrasonic water level sensors are used in the crop fields that measure
the amount of water in the irrigation channels. The sensors send their data back
via wireless transmitters.
This data is then picked up by the wireless receiver, which sends the data back to the
controller. The controller then uses this data, together with the data from the weather
station, to decide whether it is necessary to stop or start a series of water pumps. This
is done by sending signals to actuators, which operate the pumps. Although the whole
system is fully automatic, a supervisor still monitors the process remotely. Using a

schematic of a number of processes on a computer screen and via internet links to the

controllers, the supervisor can oversee several irrigation processes from one central

point. If the supervisor wishes to further increase or reduce the water supply in any of

the irrigation systems, they can override the controller if necessary.

The main advantages of this automated system are:

» reduced labour costs since the system only needs a supervisor to monitor vast

areas (if any maintenance is needed, then a dedicated team can cover all of the

irrigation systems rather than having a separate team for each system)

» better and more efficient control of the irrigation process

» better control of precious resources, such as water

The main advantages of this automated system are:

» faster response than a human having to manually check many kilometres of

irrigation channels

» safer (temperatures in the fields could be 40°C and other risks could exist)

» different crops may require different irrigation requirements (for example, rice
crops need flooding conditions, whereas orange trees like dry conditions); it is
possible to program the controllers so that different growing conditions can be
maintained simultaneously.
The main disadvantages of this automated system are:
» expensive to set up initially (expensive equipment needs to be bought)
» very high maintenance costs are associated with automated systems (also
require specialist technicians if a fault occurs, which could be a problem in some
remote areas of the world)
» increased need to maintain the water channels to ensure the system works
correctly at all times (a blocked or collapsed channel wouldn’t be picked up by the
automated system, which could result in some areas being over-watered and some
areas being starved of water).
3.Weather {Stations}

Automated weather stations are designed to save labour and to gather

information from remote regions or where constant weather data is a
requirement.
3.Weather {Stations} (Cont..)
Automated weather stations require a microprocessor, storage
(database), battery (usually with solar-powered charging) and a range of
sensors:
1. » thermometer (to measure temperature)
2. » anemometer (to measure wind speed)
3. » hygrometer (to measure humidity)
4. » barometer (to measure air pressure)
5. » level sensor (to measure rain fall)
6. » light sensor (to measure hours of daylight).
3.Weather {Stations} (Cont..)
The data from sensors is all sent to a microprocessor; any calculations
are then done (for example, calculate hours of daylight, actual rainfall
and wind direction). The data from the sensors and the calculated
values are then stored on a central database.
Some automated weather stations are sited near airports, where reports
are sent out automatically every five minutes to pilots in the vicinity of the
airport.
3.Weather {Stations} (Cont..)
The only part of the weather station that needs to use actuators is the ‘tipping

bucket rain gauge’. At a pre-determined time interval, a signal is sent from the

microprocessor to an actuator to operate a piston, which tips a bucket that was

collecting rain water. The water is tipped into a vessel where level sensors are
then used to measure the amount of rainfall that fell during the required time
interval.
4.Gaming
Gaming devices involve sensors to give a degree of realism to games:

1. » accelerometers (these measure acceleration and deceleration and

therefore measure and respond to tilting the gaming device forward/backward
and side to side)
2. » proximity sensors (used in smart touch pads; here electrodes are
embedded in touch pads that can detect hand/finger position thus increasing
user awareness).
4.Gaming (Cont..)

Embedded accelerometers and proximity sensors (together with a

microcontroller) in games consoles allow increased human interaction
with the game. This allows players to take actions that simulate real
events happening, giving a more immersive games experience.
5.Lighting
Microprocessor-controlled lighting was discussed in Section 3.1.5 (Embedded
systems).
The example used in Chapter 3 was the control of lighting in an office using:
1. » light sensors (to automatically switch lights on or off depending on the
ambient lighting)
2. » motion sensors (to automatically turn lights on in a room when somebody
enters)
3. » infrared sensors (to be used either as a motion detector or as part of the
security system).
The example we will consider here is used in a house:

1. » where lights in the garden are turned on automatically when someone

enters the garden or it turns dark
2. » where a lighting show is part of a microprocessor-controlled water fountain

display; the lighting only comes on when it becomes dark.

5.Lighting (Cont..)
As it becomes dark, the light sensor value will change, and the
microprocessor will send signals to the interface to control the array of
LED lights around the garden. Data from the infrared sensor would also
be used (during day and night) as a security device whenever the house
is unoccupied.
Also, as it becomes dark, the lighting show in the fountain could also be
initiated. This could involve a pre-programmed display involving changing
colours under the control of the microprocessor. The fountain display itself
will also be under microprocessor control with signals being sent to
actuators to turn water pumps on and off according to the installed
program. The whole system will be fully automated.
The main advantages of this automated system are:

1. » it is possible to control light sources automatically

2. » a reduced energy consumption (since lights are only turned on
when necessary)
3. » wireless connections can be chosen which are much safer (no
trailing wires)
4. » longer bulb life (due to dimming or switching off when not in use)
5. » possible to program new light displays for various occasions.
The main disadvantages of this automated system are:

1. » expensive to set the system up in the first place

2. » if wireless connections chosen (for safety reasons), they can be
less reliable than wired systems
3. » to ensure consistent performance, the automated system will
require more maintenance (which can be expensive).
6.Science

Automated systems in scientific research are widely used. There are

literally thousands of possible applications. The example we will use here
is the automatic control of a laboratory experiment which requires
accuracy and repeatability.
Imagine an experiment in a pharmaceutical laboratory where two chemicals are

reacted together in a vessel. One of the chemicals is being added from a piece of

equipment (‘A’) known as a burette (which has a tap to control the flow of liquid; the

tap is operated automatically using a small actuator) to a reaction vessel (‘B’).

Once the reaction is complete, it turns a bright orange colour (see Figure 6.11). The
whole process is under microprocessor control:
6.Science (Cont..)
The level sensors measure how much liquid is being added from ‘A’; this data is
sent to a microprocessor. Readings are also sent to the microprocessor from a
colorimeter next to vessel ‘B’ (this instrument checks the colour of the chemical
produced). The microprocessor controls the opening and closing of the tap in ‘A’;
this is done by sending signals to an actuator that operates the tap. This means
the microprocessor has automatic control of the experiment with no human
interaction needed. This type of automated system is of great benefit to
pharmaceutical companies when developing new drugs and vaccines (several
experiments can be carried out at the same time with only one person needed to
monitor the system).
The main advantages of this automated system are:

1. » more consistent (repeatable) results

2. » less dangerous (especially if the chemicals used are hazardous)
3. » faster results (several different experiments can be done
simultaneously)
4. » automatic analysis of the results is possible
5. » fewer highly trained staff needed for each experiment
6. » results/experiments can be monitored anywhere in the world in
real time.
The main disadvantages of this automated system are:

» less flexible than when using human technicians

» security risks are always present if the data is being shared globally
» equipment can be expensive to buy and set up in the first place.
Using AI in automated systems:
Finally, there are many automated systems being used in both industry and

scientific research that incorporate artificial intelligence (AI). It is therefore

worth considering the generic advantages of using AI in these automated

systems (also refer to Section 6.3):

advantages of using AI
1. » ability to access and store vast amounts of facts (very important in
research)
2. » they are able to learn from huge amounts of available data that would
overwhelm humans (or at the very least take them many months/years to do
the same analysis)
3. » they are able to see patterns in results that could be missed by humans.
The disadvantages of using AI
While all of this is positive, there are a few disadvantages in this approach:

1. » a change in skills set (is it the human or the AI that controls the research?)
2. » AI is dependent on the data which trains it.
6.2 Robotics
6.2 Robotics

6.2.1 What is robotics?

The word robot comes from the Czech word robota (which means ‘forced

labour’) and the term was first used in the 1920s play ‘Rossum’s Universal

Robots’.
6.2.1 What is robotics? (Cont..)

The concept of the robot has fired the imagination of science fiction
writers for countless years; indeed Isaac Asimov even composed his
three laws of robotics:
1. » a robot may not injure a human through action or inaction
2. » a robot must obey orders given by humans, unless it comes into
conflict with law 1
3. » a robot must protect itself, unless this conflicts with law 1.
6.2.1 What is robotics? (Cont..)

So what is a robot in the real world? Robotics is a branch of (computer)

science that brings together the design, construction and operation of
robots. Robots can be found in:
1. Factories
2. in the home
3. drones
1» factories

– welding parts together

– spray-painting panels on a car

– fitting windscreens to cars

– cutting out metal parts to a high precision

– bottling and labelling plants

– warehouses (automatic location of items)

2» in the home
– autonomous floor sweepers (see
Figure 6.13)

– autonomous lawn mower

– ironing robots (for example, ‘dressman’)

– pool cleaning

– automatic window cleaners

– entertainment (‘friend’ robots)

3» drones
– unmanned aerial vehicles (UAVs)
are drones that are either remotely
controlled or totally autonomous
using embedded systems
– can be used in reconnaissance (for
example, taking aerial photographs)
– can be used to make parcel
deliveries (for example, Amazon).
6.2.2 Characteristics of a robot

To be correctly called a robot, they need to have the following

characteristics:
1 Ability to sense their surroundings:
2 Have a degree of movement:
3 Programmable:
1 Ability to sense their surroundings:
– this is done via sensors (such as light, pressure, temperature, acoustic, and

so on)

– sensors allow a robot to recognise its immediate environment and gives

it the ability to determine things like size, shape or weight of an object, detect if
something is hot or cold, and so on; all sensor data is sent to a microprocessor or
computer.
2 Have a degree of movement:
– they can make use of wheels, cogs, pistons, gears (etc.) to carry out functions

such as turning, twisting, moving backwards/forwards, gripping or lifting

– they are mechanical structures made up of many parts (for example,

motors, hydraulic pipes, actuators and circuit boards)

– they contain many electrical components to allow them to function

– can make use of end effectors (different attachments to allow them to carry out
specific tasks such as welding, spraying, cutting or lifting).
3 Programmable:

– they have a ‘brain’ known as a controller that determines the action to

be taken to perform a certain task (the controller relies on data sent from
sensors or cameras, for example)
– controllers are programmable to allow the robots to do certain tasks.
Robotics and artificial intelligence (AI) :

It is important to realise that robotics and artificial intelligence (AI) are

almost two entirely different fields:
TWO IMPORTANT NOTES:
1 Many robots don’t possess artificial intelligence (AI) since they tend to do
repetitive tasks rather than requiring adaptive human characteristics.
2 It is important not to confuse physical robots with software robots such as:
– search engine bots or WebCrawlers (these ‘robots’ roam the internet
scanning websites, categorising them for search purposes)
– chat bots (these are programs that pop up on websites that seem to enter
some form of conversation with the web user – see Section 6.3)

According to our definition above, software robots are not true robots.
Physical robots can be classified as
» Independent robots

» Dependent robots:
» Independent robots:

– have no direct human control (they are said to be autonomous, for

example, an autonomous vehicle)
– can replace the human activity totally (no human interaction is required
for the robot to function fully).
» Dependent robots:

– have a human who is interfacing directly with the robot (the human
interface may be a computer or a control panel)
– can supplement, rather than totally replace, the human activity (for
example, in a car assembly plant where both humans and robots work
together to produce a car).
6.2.3 The role of robots and their advantages and disadvantages

We will now consider the use of robots in a number of areas, together

with the advantages and disadvantages of using robots in each of these
areas:
1. » industry
2. » transport
3. » agriculture
4. » medicine
5. » domestic (home) use
6. » entertainment.
1-Industry (The role of robots)

Robots are used in many areas of industry, from heavy lifting work right through

to delicate procedures. Examples include: paint spraying of car bodies, welding

bodywork on cars, manufacturing of microchips, manufacturing electrical goods

and automatic warehouses.

Control of robots is either through embedded (built-in) microprocessors or

directly linked to a computer system.

1-Industry (Cont..)
Programming of the robot to do a series of

tasks is generally done in two ways:

1 The robot is programmed with a sequence of instructions which allow it to

carry out the series of tasks (for example, spraying a car body with paint).

2 Alternatively, a human operator manually carries out the series of tasks; this

can be done in two ways. In our example, we will assume an object is being

painted using a robot arm.

1-Industry (Cont..)
Figure 6.18 shows a robot arm equipped with a spray gun end-effector. Different
end-effectors allow the robot arm to carry out many different tasks:

i the robot arm is guided manually by a worker when spraying the object; each
movement of the arm is stored as an instruction on the computer; or

ii the worker straps sensors to his own arm and sprays the object; each
movement is stored as a set of instructions on a computer; the sensors send
back information such as position relative to the object, arm rotation and so on –
this information forms part of the instructions stored on the computer.
1-Industry (Cont..)

Whichever method is used, once the instructions have been saved,

each series of tasks can then be carried out by a robot arm
automatically.
Each instruction will be carried out identically every time (for
example, assembling parts in a television) giving a consistent
product.
1-Industry (Cont..)
Robots are equipped with sensors
so they can gather important
information about their surroundings
and also preventing them from doing
‘stupid things’; for example, stopping
a robot spraying a car if no car is
present, or stop the spraying
operation if the supply of paint has
run out, and so on.
1-Industry (Cont..)

Robots are very good at repetitive tasks. However, if there are specialist
tasks that require ‘thinking’ to cope with variable circumstances, (for
example, making specialist glassware for some scientific work) then it is
often better to still use human operators.
Table 6.1 Advantages and disadvantages of using robots
2.Transport
Driverless vehicles are increasing in number every year. These are very complex

robots, but the big problem is not really the technology (since problems will be

ironed out through time), it is human perception. It will take a large leap of faith for

humans to ride in a driverless car or an airplane with no pilot. We are already used

to autonomous trains since these are used in many cities throughout the world.

These systems have been generally accepted; but that is probably because trains

don’t overtake other trains and have a very specific track to follow (see notes later).
Autonomous cars and buses

In this section, we will consider autonomous cars as our example.

Autonomous cars use sensors, cameras, actuators and
microprocessors (together with very complex algorithms) to carry out
their actions safely. Sensors (radar and ultrasonics) and cameras allow
the control systems in cars to perform critical functions by sensing the
dynamic conditions on a road. They act as the ‘eyes’ and ‘ears’ of the car.
Autonomous cars and buses (Cont..)
Microprocessors process the data received from cameras and sensors and
send signals to actuators to perform physical actions, such as:

» change gear

» apply the brakes

» turn the steering wheel.

Autonomous cars and buses (Cont..)

Cameras catch visual data from the surroundings, while radar and ultrasonics

allow the vehicle to build up a 3D image of its surroundings (very important

when visibility is poor, such as heavy rain, fog or at night). Suppose an

autonomous car is approaching a set of traffic lights that are showing red.

The first thing is the control system in the car needs to recognise the road
Autonomous cars and buses (Cont..)

sign and then check its database as to what action to take. Since the traffic light
shows red, the microprocessor must send signals to actuators to apply brakes
and put the gear into ‘park’. Constant monitoring must take place until the light
changes to green. When this happens, the microprocessor will again instruct
actuators to put the car into first gear, release the brakes and operate the throttle
(accelerator). This is a very complex set of operations since the microprocessor
must constantly check all sensors and cameras to ensure moving off is safe (for
example, has the car in front of it broken down or has a pedestrian started to cross
the road, and so on). To go any further is outside the scope of this book.
Autonomous trains
As mentioned earlier, autonomous (driverless) trains have been around for a
number of years in a number of large cities. As with other autonomous
vehicles,driverless trains make considerable use of sensors, cameras, actuators
and on-board computers/microprocessors. Autonomous trains make use of a
system called LiDaR (Light Detection and Ranging); LiDaR uses lasers which
build up a 3D image of the surroundings. Other sensors (such as proximity
sensors on train doors) and cameras (including infrared cameras) are all used for
various purposes to help control the train and maintain safety. The control system in
the train also makes use of global positioning satellite (GPS) technology, which
allows accurate changes in speed and direction to be calculated. Again, actuators
pay a huge role here in controlling the train’s speed, braking and the opening and
closing of the train doors.
Autonomous (unpiloted) airplanes
Airplanes have used auto-pilots for many years to control flights. Human pilots

only take over during take-off and landing. Autonomous (pilotless) airplanes

would make even more extensive use of sensors, actuators and microprocessors
to control all stages of the flight.
Autonomous (unpiloted) airplanes
a pilotless airplane would include:
» sensors to detect turbulence to ensure smooth flights
» an increase in self-testing of all circuits and systems
» sensors that would automatically detect depressurisation in the cabin; thus allowing
for quick stabilisation of the airplane
» use of GPS for navigation and speed calculations
» use of actuators to control, for example, throttle, flaps (on the wings) and the
rudder.
3.Agriculture
With the world’s population predicted to reach nine billion by the year 2050,

more efficient agriculture via increased use of robotics is inevitable. Robots

could replace slow, repetitive and dull tasks allowing farmers to concentrate

on improving production yields.

3.Agriculture (cont..)

robotics could play a big role:

» harvesting/picking of vegetables and fruit
» weed control
» phenotyping (plant growth and health)
» seed-planting and fertiliser distribution
» autonomous labour-saving devices.
1.Harvesting and picking
» robots have been designed to do this labour-intensive work; they are more

accurate (only pick ripe fruit, for example) and much faster at harvesting

» for the reasons above, this leads to higher yields and reduces waste (for

example, vegebot (Cambridge University) uses cameras to scan, for example, a

lettuce and decide whether or not it is ready to be harvested

» a second camera in vegebot (near the cutting blades) guides an arm to

remove the lettuce from its stalk with no damage.

2.Weed control
» weed management robots can distinguish between a weed and crop using AI (see
Section 6.3)

» examples of weed control robots are being used in France (by Mouton- Rothschild) to
remove weeds between grape vines in their vineyards; this saves considerably on labour
costs and improves vine growth

» weed control robots use GPS tracking to stay on course to move along the rows of vines
and remove the weeds; a weed removal blade is operated by an actuator under the control
of the controller (microprocessor) in the robot

» very often a drone (flying robot) is used first to do an aerial view of the vineyard, so that a
programmed course of action can be produced, which is then sent to the weed control
robot’s memory.
3.Phenotyping
»phenotyping is the process of observing physical characteristics of a plant in order
to assess its health and growth
» robots designed to do phenotyping are equipped with sensors (including spectral
sensors and thermal cameras) that can create a 3D image/model ofthe plant, thus
allowing it to be monitored for health and growth
» machine learning (see Section 6.3) is used to recognise any issues with leaves (for
example, if they have a blight or have the wrong colour) so that the robot can convey this
back to the farmer
» these robots are much more accurate and faster at predicting problems than when
done manually.
4.Seed-planting drones and fertiliser distribution
» drones (flying robots) can produce an aerial image of a farm sending back a
‘bird’s eye view’ of the crops and land
» they allow seed-planting to be done far more accurately
» they also allow for more efficient fertiliser-spreading to reduce waste and
improve coverage (this is much more efficient than conventional crop spraying)
» drones can also be used in cloud seeding where the drone can add silver
iodide crystals to a cloud forcing it to give up its rainwater
» the drones use a very complex camera system to target seeding and allow
fertiliser spraying.
5.Autonomous agriculture devices
Several of the devices described above could be referred to as autonomous.
The following list summarises some of the devices that can work independently
of humans:
» grass mowers/cutters
» weeding, pruning and harvesting robots
» seeding robots
» fertiliser spraying
» all of these devices use sensors and cameras to go around obstacles, or they
can even be programmed to ‘go to sleep’ if the weather turns bad.
4.Medicine
» robots are used in surgical procedures, which makes the operation safer and
also makes the procedures quicker and less costly
» robots can be used from monitoring patients to doing actual minor surgery
» the disinfecting of rooms and operating theatres can all be done by
autonomous robots (similar to the types described in agriculture)
» robots can take blood samples from patients:
– less painful to patients since the robot is better at determining a ‘good vein’
– safer to doctors and nurses if the patient has an infectious disease
– doctors and nurses can be freed up to do more skilled work
Medicine (Cont..)
» microbots can be used in target therapy:

– these use microscopic mechanical components (including microprocessor) to

localise a drug or other therapy to target a specific site causing less damage to
surrounding tissue

» prosthetic limbs are now mini robots in their own right (since they meet the three
characterisations of what defines a robot)

– bionic skins and neural implants that interface with the human nervous system
(of the damaged limb) giving feedback to allow for better control of the prosthetic
limb (again sensors and actuators are used to give humanlike responses, such as
grip).
5.Domestic robots
Robots used around the house vary from devices to carry out household chores
through to devices used to entertain people. For example:
1» autonomous vacuum cleaners:
– these use proximity sensors and cameras to avoid bumping into obstacles and
allows them to cover a whole room automatically
– these robots have a microprocessor to control the overall operation of the
device; this also allows the user to program the device
– actuators are used to control motors which allow movement forward/ backward and
from side to side
Domestic robots (Cont..)
2» autonomous grass cutters (mowers):
– these use the same type of sensor, camera, microprocessor and actuator set up
as vacuum cleaners
3» personal assistants (such as ‘Vector’)
– this is a robot controlled by a micro-processor that also uses cloud connectivity
to connect to the internet
– it understands voice commands (using a microphone) and will answer any
questions it is asked
– it also makes use of an HD camera, utilising computer vision, allowing it to
recognise somebody’s face as well as navigate a room (using proximity sensors
and actuators) to steer around objects in its way.
6.Robots used in entertainment

The use of robots in the entertainment industry is increasing. They are

now found in areas such as:
» entertainment parks and arenas/venues
» the film and TV industry.
Robots used in entertainment (Cont..)
The following examples indicate where robots are being used in the world of

entertainment. The reader is advised to research the ever-increasing number of

examples:

» theme parks are now using autonomous robots to entertain visitors to the

park; these robots (often dressed as cartoon characters) can interact with

visitors to allow them to engage safely with the theme park attractions and

make the whole experience ‘more realistic’

Robots used in entertainment (Cont..)
» music festivals are much more immersive for the audience; robotic methods
are used to control lighting (including laser displays), visual effects and
animation (e.g. superimposing an actor’s image onto a robotic caricature and
synchronising mouth movements); the visual performances can be fully
synchronised with the music

» use of robots to control cameras; for example, keeping them steady and
autofocusing when moving around a scene; the movie Gravity used many robots
to operate cameras, props and the actors (for example, to give an actor the
appearance of moving around in the vacuum of space uncontrollably, robot arms
were used to simulate human behaviour and produce life-like moving images)
Robots used in entertainment (Cont..)
» humanoid robots (either remote-controlled or pre-programmed) can perform
‘stunt’ action in movies/ television by performing tasks impossible for a human
to do; they use CGI (computer-generated imagery) and image capture techniques
to generate special effects

» robots are capable of producing special effects with a precision, speed and
coordination which is beyond human capabilities; actions and special effects
can be synchronised to within a millisecond and produce fully
coordinated/synchronised sound effects (e.g. movement of the mouth to match the
sounds produced in a realistic manner).
6.3 Artificial intelligence (AI)
6.3 Artificial intelligence (AI)
6.3.1 Introduction
Artificial intelligence (AI) is a branch of computer science dealing with
the simulation of intelligent human behaviour by a computer.
This is often referred to as the cognitive functions of the human brain
(that is, the mental process of acquiring knowledge and understanding
through thought, experience and the five senses). All of these cognitive
functions can be replicated in a machine, and they can be measured
against human benchmarks such as reasoning, speech and sight.
6.3.2 Characteristics of AI
Essentially, AI is really just a collection of rules and data, and the ability to reason,
learn and adapt to external stimuli. AI can be split into three categories:

1» narrow AI – this occurs when a machine has superior performance to a human

when doing one specific task

2» general AI – this occurs when a machine is similar (not superior) in its

performance to a human doing a specific task

3» strong AI – this occurs when a machine has superior performance to a human

in many tasks.
6.3.2 Characteristics of AI (Cont..)
Reasoning: is the ability to draw reasoned conclusions based on given data/

situations.

Deductive reasoning is where a number of correct facts are built up to

form a set of rules which can then be applied to other problems.

(for example, if AI is used to produce the perfect cup of tea based on a number of
facts, the machine will learn from the experience and apply its new rules in the
making of a cup of coffee, hot chocolate, and so on – modifying its methodology
where necessary).
6.3.2 Characteristics of AI (Cont..)

By carrying out a sequence of steps, the AI machine can learn, and next
time it will know how to do the task more effectively and even apply it to a
novel/new situation. Thus the AI system is capable of learning and
adapting to its surroundings. AI can very quickly discern patterns (which
in some cases, humans cannot) and then make predictions by adapting to
the new data. How all this is done is beyond the scope of this book
(interested readers can find out more about this topic by consulting
Cambridge International AS and A Level Computer Science, ISBN:
9781510457584).
Examples of AI include:
1» news generation based on live news feeds

2» smart home devices (such as Amazon Alexa, Google Now, Apple Siri and
Microsoft Cortana):

– the AI device interacts with a human by recognising verbal commands

– it learns from its environment and the data it receives

– the device becomes increasingly sophisticated in its responses, thus

showing the ability to use automated repetitive learning
Examples of AI include:

3» use of chatbots that interact through instant messaging, artificially

replicating patterns of human interactions using AI to respond to typed or
voice messages; when a question is asked, the chatbot responds using
the information known at the time:
Examples of AI include:
4» autonomous cars (see Section 6.2)

5» facial expression recognition:

– algorithms identify key facial landmarks such as the corners of the eyebrows,
corners of the mouth, and so on

– a combination of these landmarks can be used to map emotions (such as

anger, fear, joy and surprise).

6.3.3 AI systems
This section considers two types of AI system:

1» expert system – a computer system that mimics the decision-making ability of

a human; expert systems use AI to simulate the judgement and behaviour of a

human or organisation that has expert knowledge and experience

2» machine learning – this is the science of training computers with sample data

so that they can go on to make predictions about new unseen data, without

the need to specifically program them for the new data.

1.Expert systems

Expert systems are a form of AI that has been developed to mimic

human knowledge and experiences. They use knowledge and
inference to solve problems or answer questions that would normally
require a human expert.
Expert systems (Cont..)
For example, suppose the user was investigating a series of symptoms in a

patient. The expert system would ask a series of questions, and the answers

would lead to its diagnosis. The expert system would explain its reasoning with

a statement such as ‘impaired vision, lack of coordination, weak muscles, slurred

speech and the patient used to work in a paint factory – the diagnosis is mercury

poisoning’ – the user could then probe deeper if necessary.

Expert systems (Cont..)

The expert system will supply a conclusion and any suggested actions
to take and it will also give the percentage probability of the accuracy of
its conclusions (for example, the following statement could be made
‘Based on the information given to me, the probability of finding oil
bearing rocks in location 123AD21G is about 21%’).
There are many applications that use expert systems:

1. » oil and mineral prospecting

2. » diagnosis of a patient’s illness
3. » fault diagnostics in mechanical and electronic equipment
4. » tax and financial calculations
5. » strategy games, such as chess
6. » logistics (efficient routing of parcel deliveries)
7. » identification of plants, animals and chemical/biological compounds.
Expert systems have many advantages:
1. » they offer a high level of expertise
2. » they offer high accuracy
3. » the results are consistent
4. » they have the ability to store vast amounts of ideas and facts
5. » they can make traceable logical solutions and diagnostics
6. » it is possible for an expert system to have multiple expertise
7. » they have very fast response times (much quicker than a human
expert)
8. » they provide unbiased reporting and analysis of the facts
9. » they indicate the probability of any suggested solution being
correct.
Expert systems also have disadvantages:

1. » users of the expert system need considerable training in its use to

ensure the system is being used correctly
2. » the set up and maintenance costs are very high
3. » they tend to give very ‘cold’ responses that may not be
appropriate in certain medical situations
4. » they are only as good as the information/facts entered into the
system
5. » users sometimes make the very dangerous assumption that they
are infallible.
Expert system structure
So what makes up an expert system? Figure 6.23 shows the typical structure of
an expert system:
User interface
» method by which the expert system interacts with a user

» interaction can be through dialogue boxes, command prompts or other input

methods

» the questions being asked usually only have Yes/No answers and are based
on the responses to previous questions.
Inference engine
» this is the main processing element of the expert system

» the inference engine acts like a search engine examining the knowledge base for
information/data that matches the queries

» it is responsible for gathering information from the user by asking a series of questions
and applying responses where necessary; each question being asked is based on the
previous responses

» the inference engine is the problem-solving part of the expert system that makes use of
inference rules in the rules base

» since the knowledge base is a collection of objects and attributes, the inference engine
attempts to use information gathered from the user to find an object that matches (making
use of the rules base to find a match)
Knowledge base
» the knowledge base is a repository of facts

» it stores all the knowledge about an area of expertise obtained from a number

of expert resources

» it is basically a collection of objects and their attributes; for example:

Knowledge base (Cont..)
» so if we had a series of questions:
– is it a mammal? YES
– can it be a pet? NO
– does it live in water? YES
– does it make sonic sounds? YES
– is its body covered in skin? YES
– does it have any legs? NO
conclusion: it is a WHALE.
Rules base
» the rules base is a set of inference rules
» inference rules are used by the inference engine to draw conclusions
(the methods used closely follow human reasoning)
» they follow logical thinking like the example above; usually involving a
series of ‘IF’ statements, for example:
IF continent = “South America” AND language = “Portuguese” THEN country = “Brazil”
Setting up an expert system
» information needs to be gathered from human experts or from written sources such as
textbooks, research papers or the internet
» information gathered is used to populate the knowledge base that needs to be first created
» a rules base needs to be created; this is made up of a series of inference rules so that the
inference engine can draw conclusions
» the inference engine itself needs to be set up; it is a complex system since it is the main
processing element making reasoned conclusions from data in the knowledge base
» the user interface needs to be developed to allow the user and the expert system to
communicate
» once the system is set up, it needs to be fully tested; this is done by running the system with
known outcomes so that results can be compared and any changes to the expert system made.
2.Machine learning
2.Machine learning (Cont..)

Machine learning is a sub-set of artificial intelligence (AI), in which

algorithms are ‘trained’ and learn from their past experiences and
examples. It is possible for the system to make predictions or even take
decisions based on previous scenarios.
2.Machine learning (Cont..)
They can offer fast and accurate outcomes due to very powerful

processing capability. One of the key factors is the ability to manage and
analyse considerable volumes of complex data; some of the tasks would take
humans years to complete without the help of machine learning techniques. One
example that uses machine learning are the most sophisticated search engines:

The search engine will learn from its past performance, meaning its ability to

carry out searches becomes more and more sophisticated and accurate.
Consider email messages such as ‘You have won $2 million in the National Lottery’; how can
machine learning determine that this email should be put into your spam folder?
» A machine learning algorithm collects data about emails, such as email content, headers,
senders name/email address and so on.
» It carries out a ‘cleaning’ process by removing stop words (for example, the, and, a) and
punctuation, leaving only the relevant data.
» Certain words/phrases are frequently used in spam (for example, lottery, earn, full-refund) and
indicate that the incoming email is very likely to be spam.
» The machine learning model is built and a ‘training data set’ is used to train the model and make it
learn using past email known to be spam.
» Once it is evaluated, the model is fine-tuned and tested live.
When you visit an online retailer, such as Amazon, you might receive the message
‘customers who bought Hodder Education IGCSE ICT textbook also bought Hodder
Education IGCSE Computer Science textbook’. How is machine learning used to
establish a user’s buying characteristics?

» This comes from collaboration filtering, which is the process of comparing

customers who have similar shopping behaviour to a new customer who has similar
shopping behaviour.

» Suppose customer ‘A’ is very interested in playing football and they also bought a

jazz CD, a book on Roman history and some health food.

» Two weeks later, customer ‘B’ who likes to go cycling also bought a similar jazz CD

and a book on ancient Roman history.

» The machine learning algorithms will then recommend that customer ‘B’ might

like to buy some health food due to the similarities between ‘A’ and ‘B’s shopping

behaviour.

» This technique is particularly popular when asking your mobile phone to generate

a playlist from your music library based on a few criteria you might select.
» Data is gathered by a survey or web scraping; for example, to detect credit card

fraud, information about customers is gathered, such as types of transactions (does

the customer buy designer clothes?), shopping habits and certain personal data.

» Redundant data is then removed; this needs to be carefully done to remove the

possibility of wrong predictions.

» The most important machine learning step: the algorithm is trained through real

examples of customer purchasing behaviour.

» A model is built based on learning from the training data, and the machine

learning algorithm can now be used to detect fraud (for example, if a customer

spends an unusual amount on a piece of jewellery, there is a high chance a

fraudulent activity has taken place).

» The machine learning model is then fully tested with known data and known

outcomes; the system is modified if it hasn’t met its criteria to detect fraudulent

activity.