Appunti per l'esame di Digital manufacturing

Beyond those technologies there are 2 main trends that enabled industry 4.0 that are:

- Moore’s law: evolution of computing power, as the number of transistor that could be fitted in a

processor doubles almost every year;

- Networks: needed to transfer data collected by sensors.

So, companies must deal with many technologies that today are all pretty much mature. From an

organisational perspective company experience the so-called combinatorial e ect of the

technologies. This means that if I adopt a technology, I have an improvement, then, if I adopt 2

technologies I have a further improvement, but it doesn’t double, the e ect is a combination of the

e ects of both technologies. The use of di erent technologies implies the combinatorial e ect, for

which the use of di erent technologies together has a higher benefit than the sum of single benefits.

Impact of digital transformation on the factory = can be explained with the concept of the Smart

Factory that is the evolution of traditional production lines with automated systems. Having CPS

(autonomous collaborating entities able to act as physical objects and communicate on global

networks, compute and store data in the cyber world through their computerized version named digital

twins) in the factory means having more information, improving the quality and speed of decisions. In

this type of systems, any entity involved, tangible and intangible ones, have a corresponding digital twin,

otherwise the CPS cannot be established. To extract data used to generate a digital twin there is a

second fundamental factor that are Big Data, as data must be considered one of the inputs together

with raw materials and workers. Data if properly used can influence company performance by improving

the productivity of the machines. Another important term to be introduced is Automation 4.0,

determined by these historical steps:

- Industrial automation: replace workers with robots (70s robots automatized);

- Collaborative automation: collaborative robots work together with people;

- Cognitive automation: data given to a human decision maker who uses them to make

decisions. Workers decide things instead of doing stu .

Couple of issue in a flat manufacturing system

In a flat manufacturing system the North-South tra ic in datacenters is not present anymore, but there

is an horizontal tra ic, the machines are connected between them. Indeed, the Smart Factory changes

the way in the information systems of a company, there was a switch from MES to MOS. They both are

information system that allows to control and manage the manufacturing of a company. The main

di erence between them is that MES (Manufacturing Execution System) collects data on a unique

centralized DB, while MOS (Manufacturing Operating System) enables machines to have all information

inside them, they keep data close to the machine that generated them. So, in the past, when a machine

had some problems, it had to share the problem from the device level on the up level, then the signal

went down to the device level. In the future, the equipment will send the request directly to the

equipment of the supplier.

Further component that enables Industry 4.0 are:

- Lean Manufacturing4.0 = optimising the processes to improve their e ectiveness and

productivity. The base idea is that it’s not useful to collect data if we don’t use that data to

improve the processe because we still have processes that are not e icient and we can’t reduce

waste. Not doing this step means that we are only digitalising;

- Service 4.0 = application of disruptive technologies in services, digitalization of a service in

order to improve the e iciency. To introduce it we can make the example of jewels shops in

Paris, which placed cameras in the windows to react potential customers face expressions

watching the window. The reactions are elaborated by a sentiment analysis algorithm that is

able to associate the person, the jewel and the reaction, this way the shop manager knows

which jewels o er to the clients. This results in an increase of the sales probability.

Industry 4.0 increases productivity = nowadays there are a lot of people that are complaining about

the use of AI that could reduce the #workers in a company, but it is not true because the productivity of

a company is made of 3 factors that are work, tech and data. So, in this case when there is more tech,

there are less people, but then you become more competitive, and you need to hire more specialized

people because you sell more and you need people with more competences. If you leverage more

productive data, more productive tech. Also, if there is an increase in productive of data, you have an

increase in productivity of the tech.

2-INDUSTRIAL MEASUREMENTS

In order to have a fully automated system we need sensors to take measures and collect data. To

handle the measurements to control the fully automated system we can adopt the traditional

measurement chain.

Traditional Measurement Chain = sequence of devices (or steps) that we must do in order to get and

then send a signal, the devices are:

 Sensor = reacts to a change on a variable that we want to measure;

 Transducer = transform the reaction of the sensor into an electrical signal;

 Transmission cable = any kind of mean that is able to transmit the signal, like cables or wi-fi;

 Conditioner = takes the electric signal created by the transducer and make some

modifications to improve it. Usually it can amplify, reduce or filter to obtain the desired

bandwidth;

 Converter = makes visible the signal, so that converts a continues signal into a digital one;

 Computer = shows the signal.

Sensors can be:

- Active: have an internal source of power to work;

- Passive: requires external source of power to work.

Integrated Sensors = sensors that integrates the sensor, the transducer and the conditioner. Some

advantages are the no need of calibration (adjustment according to a specific scale) and the reduction

of the assembly cost.

Smart sensors = sensors that have some functions inside like the ability to take a decision based on

measured signal, to transmit and receive signals, the auto-calibration, diagnosis of malfunction and

the standard connectivity.

Thermocouple sensors = sensor that measures the temperature leveraging on the Seebeck e ect,

that is a physical phenomena which is happening when we’re using 2 di erent metals. There are two

points, the hot point that we want to measure and the cold point. If there is a current that passes

between the points we are able to measure the di erence of potential between the 2 connected parts

when there’s a di erent temperature. It’s important to highlight that this is a di erence, so we have to

do something to measure the temperature of the hot point. For this reason, the Wheatstone bridge

was invented, an electrical structure that can compensate the variation of the temperature in the cold

point, in the case it decreases. So, the Wheatstone bridge takes action and thanks to this circuit we

are able to keep the cold point fixed and measure the temperature of the hot point.

Pressure sensors = sensors that transforms electric signals in pressure. There are 3 di erent

typologies:

- Electrical extensometer: there is a tank with a predefined pressure and with a hole inside,

then there is a screw with a resistance inside. When we put the screw in the tank, there is an

extension of the resistance for the pressure present in the tank, there is a current that passes

in the screw so we can understand the di erence of potential and then with a formula even the

pressure;

- Piezo resistive: equal concept but with a di erent material, silicon chip instead of ceramic.

- Capacitor transducer: there are two pieces of metal close one each other, when they have

been interacted with a pressure they receive a di erent extension and become closer, so there

is a di erent electricity potential that we can use to measure the pressure.

Position sensors = sensor to measure the distance from an object. It is characteristic of two lens

placed between the object they reflect the light to a specific area and with goniometry with the angle

we can understand the distance of the object. The are receiver of the light is placed in a diagonal way

so the area is wider.

Speed sensors = used to measure the speed like

- Positioning sensor: as speed is space divided by time we can make two di erent

measurement and then divide;

- Tachymetric Dinamo = a mechanical equipment that rotates and transform the mechanical

energy into power. Higher are the rotations per minute, higher is the produced electric voltage

according to the formula: N(#rotations)*flow;

- Impulse wheel = a rotating wheel in which there are some teeth and when they pass the root

of the sensor they generate an impulse, the system measures the frequency of the impulses,

the faster is the rotation of the wheel the highest is the frequency measured. According to the

number of impulse received we can understand the speed.

Flow sensors = they consist of a pipe and a tube to measure the flow of a fluid in a tube, with some

fluid flowing into the pipe. The sensor has a turbine that rotates accordingly to the flow, so if the flow is

higher the turbine rotates faster. The mechanism is the same as before, every rotation of the turbine

generates an impulse. In some cases we can have a very small flow, it’s a problem as the turbine

cannot move if the flow is too low. We can change the design of the turbine rotating it.

Level sensors = there are two types of sensors. The first one gives a binary result, the second one no.

The first one is made of a led light that it is placed in a plastic cone that generates a led light. If this

light is reflected it means that the plastic cone reached the level of the fluid. In this way you can

understand if the object is at the level of the fluid. This solution gives only a banal response that it is 1

or 0. To overcome this problem there is a solution that has a glass cylinder in which there is the fluid

and there are 2 floating balls, that move according to the level of the fluid. The upper ball magnetized

to understand the level of the fluid because there is the movement of the magnet in the ball since

there is series of magnets along the glass cylinder that magnetizes the first floating ball. The bottom

ball is used to float the upper one.

3-PROCESS CONTROL

Process Control Logic Schema = the main module is the production process that receives in input

the raw materials and that gives in

output the finished products. During

this process we can collect 5

di erent types of data that are:

1.Raw materials status: data about

quality and quantity of raw

materials;

2.Process variables: variables that

a ect the finished products like

temperature, capacity and

composition;

3.Data related to equipment:

physical infrastructure where the

process is happening, if there are

failures;

4.Product quality: read data from

the output, like the number of

defects;

5. Environmental con