A report on Industrial Automation

Taavi Klaos attended my lectures on Electronics and Communication Industries
(IED3100) at TTU - Tallinn University of Technology, Thomas Johann Seebeck Department of Electronics. At the end of the course he wrote a paper on Industrial Automation that I would like to introduce in the following lines.

Today we have in industry automation three main types: PLCs(programmable logic controllers), PACs (programmable automation controllers) and IPCs (industrial programmable computers). Most common of them is PLC, beacuse it is oldest, cheapest and easiest way to automate production or application. PAC and IPC are used, when there is more complicated activity or it needs some special addition. 

On last decades difference between PLCs, PACs and IPCs have decreased because PLCs have obtain capabilities of motion control, advanced proportional-integral-derivative PID process control, integrated safety, features like as a web server and networking utilities. Generally PLCs and PACs are equalizing and only IPC has a bigger difference. At the moment PLCs and PACs are different in programming language, which is the biggest difference. PLCs are usually using ladder logic, but PACs are developed with more generic software tools.

Picture 1 Siemens PLC Picture 2 Relay automation

Industrial automation with PLCs came in the early 1970. Before that all the logic gates were done with relays and it took a lot of space and it was quite complicated. Relays had to maintain every year for durability and it took time and was expensive. In Estonia I have seen until today relay based automation. There were hundreds of relays clicking. I had to understand why one function wasn´t done and it was really difficult without drawings. First commercial producer was Modicon, after that came Allan Bradley and other companies. 

Communication abilities appeared in 1973 and was called Modicon´s Modbus. This gave possibility to communicate between PCs and PLCs. Next decade was nightmare for communications, because there wasn´t standardization and technology was changing. At the moment some biggest manufactures of PLCs are Siemens, Mitsubishi and Rockwell. 

Every continent has his own main producers and by this we have a lot of different communication protocols and standards. In Europe main producer is Siemens , which gives us fieldbus Profibus. Only the main brands are all over the world. Also it is easier to use products like Siemens, because there is a lot of information about Siemens programming and other help. Minus side is higher price so hardware than software. Newcomers frequently give software for free and try to sell more hardware like this. The first who was doing in Estonia this strategy was Unitronics and achieved good position. Also they had really good  software suppert.

Last bigger announcement was PAC(programmable automation controller) and it was first defined by ARC Advisory Group in 2001. Best way to introduce PAC is something like this: Programmable Automation Controllers that combine the functionality of a PC and reliability of a PLC. PAC is a 2 or more processor based device like a personal computer(PC), but orientated for industry automation. At the beginning PACs function were doing IPCs. This were too expensive and for that was developed controllers, what fits between PLCs and IPCs. Some producer called they high end PLC products for PACs. End user can´t normally see difference of them and call both of them PLC.

Picture 3  Siemens IPCs

Industrial PCs (IPCs) are very resistant computers for environment, because thay are used in mainly in hard conditins. They came a little bit after personal computers on the 1980s. An industrial PC is PC-based computing platform for industrial applications, processor is the same as everyday computers have. They even look like usual competers in harded case. For touchscreen solutions there are IPCs like screens and back of it there is processor with communication interfaces. Robustness  is achieved with heavier metal construction as compared to the typical office non-rugged computer, additional cooling counting with industry speciality, electronic components may be selected for their ability withstand higher and lower operating temperatures. 

IPC could do many different applications in the same time and process large amounts of data. Last types of IPC has same hardware as personal computers and with this there isn´t a big hardware development costs.

Markets like PLCs are saturated and there is a very little room for newcomers however IPCs world is splintered by little specialized companies . They usually come from Germany or Tawain and have a main district of applications. German suppliers typically offers modular and open IPC hardware for specific solution every industry then Tawainese providers started as original design manufactureres and now also offer full-blown IPCs. In Europe in Siemens one of the biggest IPC producers and they have good range of different IPCs. I see this is very little market compared with PLC and for newcomers it is easier to survive. 

Main thing is to choose one application and concentrate on this.

Main advantages of IPCs are that industries need to handle ever-growing amounts of data. This data is often used for feedback to optimize production or to fulfil regulatory requierements for collecting and storing data. In these days all data is located locally, but in future this is going to cloud. This mean that  good and fast internet connection in very important, meanwhile there should be backup near to you. This will give mirrored data and security for data lost. Most of IPCs are based on windows operating systems, what is easy to use for all computer users.

In the last decade we have automation everywhere in industries and without this we can´t image a factory. PLCs are widely spread and it is very easy to use. Cheapest ones cost below 100€ and software is free, what gives demand for all kind of applications. Also you could expand them and add different modules. PLCs are replacing all logic gates with realibility and flexibility. For consumer it is easier to buy controller and use it in many ways than do something difficult with many little things like program clock and relays.

In the future PLCs could adopt more improvments specially from communication field. Programming and connection should go over RJ45 socket, what is already happening. Older ones use RS-232, RS-485, USB or something special from PLC producer. 

There are converters for RS-232 to USB and so on, but this makes sometimes problems and you need to adjust computer for this. In some conditions it could freeze computer or even do restart, I have feel it with mine. When we have identical plugs and sockets on controllers we can think about communication standards. Main hope from consumers is that producers will obtain one standard and use it. At least not more than four five standards. At the moment we have Profibus, Modbus, RAPIEnet and lot of other automation communication standards. After that there is different standards for building automation too, called KNX, EIB and so on. With this we could save much money from converting different signals. 

Converting signals also causes problems with constant data flow, because converter could lose some data. At he moment we have a lot different programming cables and standards and this is very costly and gives headaches. Even one produces will change the cable standard after some years and by that programmers should have a lot of them. Better programmers have a little suitcase for different cables, what they use communication with controllers. For programmers are computer manufactures even doing special computers with different interfaces. Good example is Siemens, who is selling ready to use. Robust programming device in notebook look for automation engineers.

Programming languages like in ladder logic and functionblock could be the same for next decade, because this is easy to use for everyone. PLCs should be the easiest way to automate machines and programming language is the main thing in this. For PACs and IPCs there are more difficult languages such as C or C++ and an off-the-shelf package such as TwinCAT and Wonder Ware. I can´t see for PLCs, PACs or IPCs better solutions for programming on the next decade. Maybe the easiest programs should come with software and could be adjusted  for specific work. This is commen  in building automation controllers, but in industrial PLCs I haven´t seen it.

Second bigger thing is hardware. At the moment we could see on the market first PLCs and PACs based on mobile phone or tablet hardware, what is widely spread and give better prices. These solutions uses mainly same parts of your mobile phone and will be easy to handle. Also they are more efficien than usual PLCs. This will give for the hardware really good price and open software like android. It is realy good to use touchscreen, because we have allready tablets 10 inches even bigger. In Estonia we can see first company, who is producing these controllers with based android software. They are advertising this with these sentences. Android based PLC has minimum 24 input and output and it could be larger. It could me programmed in many languages. Price is low and could used in very different applications like traffic light-, street light-, pumping station-, smart building managing. In the future we should see on the consumer electronics market with open eyes and used them on other applications.

I see more adoptions from widely spread consumer technology. One region is gaming consoles and additions of this. Latest technology of gaming consoles what is used more and more is 3D camera, what is used in game consoles. It game first with Nintendo Wii and has it Microsoft Xbox and others too. Companies are selling this solutions in ten millions and price is going down. I see 3D future in automation industry as monitoring and managing. Operator could see and change different views by waving his hands. Operator does not need with this a lot of buttons on the board or touch screen, which is bigger than him. All the necessary information is displayed on the air or on the wall and camers could see where operator is pointing. It seems like from movie about future, but it is next bigger thing of industury automation supervisory and managing.. Only for the industry it is very expensive too develop new hardware or future solutions, but adoption from other regions is cheaper as development costs are lower.

Speaking more about communication and data flow we can´t forget Wi-Fi. Very interesting is to see how will spread Wi-Fi in industry automation. It could be one part of controllers and sensors communication way. It is not possible to get good connection from far-away places or through many walls, but factories are usually with big open rooms and  there signal will spread good. In everyday life we have almost in every building Wi-Fi and why not to use this for communcation in factories. It could not used in very fast processing where speed is very important, but in usual data transfer is possible. Fiber-optic cables goes to main routers, what should have 802.11ac standard which gives speed up to 1300Mbit/s total. We can´t use usual Cat5 or Cat6 cables, because these maximum speed is lower. 

In the future Wi-Fi speed could be even faster. Scientist are done it, but it is not eady for commercial use. With this there isn´t a problem of speed or amount of data. For controllers this mean that we need to add Wi-Fi receiver from hardware and do software changes. With signal sensors it is much more complicated, because these should be compact as possible. One possibility is to collect for example 4 sensors output and have there a reciever and transmitter. Solutions like that will but a lot of signals in the air. Advantage is less work for cable installers and simpler structure. Is these signals affecting humans I don´t know. In everyday life we live apartments and  there are sometimes about twenty different transmitters and recievers.

It is common that controller main unit have few inputs and outputs(I/O) attached. Usually these are digital I/O and for analog there you need to buy addition module or when ever you want to expand you controller. Additional signal modules normally divides for two: analog and digital signal additions. For the future I see that controllers have only digital I/Os. Analog signals for example 0-10V are very sensitive for noise what comes from power cables. It could make troubles for reading measurement and make cable mounting price higer. You need to but cables futher away from AC power cables, have a really good shield isolation, voltage signal cables can´t be longer than 20m(not accurate with longer cable)  and even then the noise could have a effect. There is a great possibilty with voltage sensors that power cables does inductive voltage inside it. 

Voltage sensors are used mainly in building automation or low cost solutions. In industries are mainly used 4-20mA signal sensors. This does not admit so much disturbance and could used over 200m long cables, but still we have not cleared noise or some affectence. For that I see in the future that sensors are giving out digital signals. This mean that sensors costs more and have more electronics like analog digital converter. Also it means that we have a one common standard for sensors communication. With this we can see more accurate measures, cheaper and easier controllers. Other solutions is to have little data collecting points, where are analog signals collected and converter to digital signals and sent to controller. Solutions like this is already used and it is working.

Picture 4  Industrial robots Comau

In the next years the biggest profit making market in industries for installers is industrial robot and their maintenance. This market came on the eighties and is used in very big factories. In the world there is few producers of industrial robots and products are expensive. In the Europe we know some bigger manufactures like KUKA, ABB and Comau. Largest markets for robots are in Europe Germany and Italy. Only country whos market in industrial robots is every last year rising is Czech. 

Adjusting industrial robot to work is difficult and needs experience. It is not compared with PLCs programming or something similar. Industrial robots have a working body and hardware near to body. Industrial robots remind us industrial programmable computer with body attached with it. Programming industrial robot could be done with two ways. One option is to programm it by computer, but you need to know very accurate dimensions of movements. The second way and more common is done this by pendant. With this you could move robot diferents parts of it and save moves, points and adjust  the speed. Usually installer chooses points of movement way with pendant and after that finish programm in computer. Computer interface is better to choose different speed and approaching angles.

People who has a knowledge could restore old industrial robots, because these have a good value even after restore. Bigger companies often buy new robots rather than will do a big service. Car factories buy hundreds of robots at once an will change them with same way. After that could used robots restore in smaller companies and used again. It is very usefull for smaller and not so rich companies. In Estonia industrial robots aren´t very common, but we have jobs what could be replaced with robots. With this we can get rid of easy jobs and be more innovative. Also with robots we need more intelligent people and  pay higher salaries to give service for robots.  

Supervisory system are spreading more and more in industries. It gives overview of factory and helps to see barriers on factory line or some other errors. Main option of supervisory is data and error management. It is very important to get error messages on the same moment as this occurs. Usually this is done by notifications from gsm, email or alert on computer screen. When we have a notification we could resolve it and everything would work again. Other important thing is data and command management. Operator should see everything, what is happening in the plant. He is like a captain in the ship. Important is even add utilities attached to system like electricity and water. Better is to have utility measurements from many places of factory area. 

With this we could see how much different parts are spending electrity and water and do a good survy of costs. With good supervisory system operator could avoid stops and brakes. Other thing is that this needs to train workers who can see even a little difference and how to react with different occasions.

Picture 5  Google glasses

One possibility how to make industrial automation installer life easier is in the future use smart glasses. At the moment this is only in developing position for public consumer, but when it goes popular we  can see it on other applications too. This could give us handsfree picture from different sections in factory and online reports. One option is phone or tablet, but this is usually in you pocket or shoulder bag and you could not notice. In bigger factories it could help operator to solve problems. Smart glasses should have factory plan and when problem notice comes, it directs you to right place. 

Industrial automation is basically not too complicated. It goes complex with size of application and supervisory. Doing a program with PLCs with gsm feedback is basic, but when you add visualization or different communicatoin ways it needs more time. Today we can´t have a industry without automation. From that we can see industrial automation is growing market and it needs already to renew ten or twenty years oled programs. Second thing is that we need to solve with automation continuosly more difficult application. This gives to our factories more robots and sequence programs. In Estonia industrial automation came on tha late nineties. This mean for Estonian industrial automation installers more job from old arrangements.