Automation in Everyday Life
Published by : Industrial Automation
This is the first in a series of articles by Sivaram PV, sketching out what the articles in this series would examine.
Control challenges in production of objects of everyday use are many. A close look at the kind of challenges that are encountered opens an interesting puzzle. Problem categories are few; because the object parameters which need to be controlled are few. These could be – for example – physical dimensions of the object, weight, thickness, surface finish, etc.
Corresponding control variables are nearly always the same – speed, rate, tension, position, thickness, pressure, level, temperature and so on. Control algorithms themselves are notably limited – proportional control, integral control, PID control, etc. Where is the puzzle we mentioned at the beginning of the paragraph? The puzzle is so many challenges that arise during implementation in an industry situation. These challenges arise out of non-ideal measurements, control responses, imperfect earthing and similar issues. In spite of all these shortcomings, the machine must function properly and deliver good products. This is the magic achieved by clever control algorithms. Control schemes and algorithms are the solution to the puzzle.
Laboratory and industry
This series is meant for students; but even more for faculty. There is always a gap between learning theory from textbook or class room, and learning from a hands-on lab experiment. The lab experiment is oriented to deepen understanding of theory. Too frequently one encounters an opinion that academic design is not useful in industry practice, and there is a gap between the lab and the shop floor. This series of articles expresses the view from industry, and hopes to bridge this gap. Each article can be a starting point for creating a lab setup for investigating actual issues in industrial machines.
A disclaimer – we do not delve into theoretical aspects of control loops and control design. The story-line is from industrial implementation point of view.
Control and automation and data technologies
Automation is a topic of Industrial Revolution THREE, and can also be called Digitisation. Data technologies – which are the backbone of Industrial Revolution FOUR – are means to monitor and assist managers to make competent decisions, based on events and measurements from the shopfloor.
Control is the effort is the design effort to keep a chosen parameter within defined limits, even as the process environment is changing. Control predates digitisation and even electronic controls. Automation is a means of defining in great detail a control process, and a means to perform the control task tirelessly and repeatedly according to the programming.
Automation is technically not a necessary element for data technologies. Yet, if the elementary parameters of throughput and quality are not achieved using good control automation, benefits of data techniques cannot be obtained.
Basics of control loop
There are typical elements in a control loop. There is first and foremost the element to be controlled – it is called control variable. It could be level of liquid in a tank, it could be temperature of a substance, or similar. Next there is a set point. This is the desired value to be attained or maintained by the control variable. There will be a measurement of the actual value of the control variable – what is its value now. The difference between set point and control value is the deviation. This deviation will act on a control element – a valve or a heater or something, in a manner to reduce this deviation.
We will explore many control schemes. Every month we will take up some objects which we encounter in everyday life. We will touch briefly on some processes in manufacture of these items. We will sketch out the control schemes employed to improve productivity in this manufacture. We will give some hints as a starting point for faculty and students to device experiments in controls using automation.
An efficient control algorithm is one which keeps the process variable (PV) within permitted limits around the set-point. More the number of excursions, higher the amplitude of the excursions, worse is the algorithm. Some of the factors which influence behaviour of a control algorithm are – Inertia of the load system, noise in measurement, frequency of sampling of control variable, rate of change of process variable (jerk change for example), rate of execution of the control program itself and so on.
Control algorithms and PLC programs
A PLC program is representation of the control algorithm in software. A typical PLC program cycle works like this – Initially all inputs are scanned and recorded in the local memory. This constitutes the input image, and is assumed consistent, that is, all values have same time-stamp. Then the logic is executed, and outputs are calculated and written to output image. Finally the output image is transferred to the actuators. Hence one can see that there is, in worst-case, a latency of 3 cycle times. The discussion becomes more complex, if you consider remote IO systems, multiple controllers communicating on a bus, and what is very common – control loops inside control loops.
Inertia of load is not always mechanical inertia in moving parts. It is sometimes for example thermal inertia, like when you try to heat an object, depending on its heat capacity, response (increase in temperature) can be slow. The control algorithm needs to have this inertia and the latencies as factors.
Control algorithms need tuning for a given installation. This is mostly a manual input, and provided by experienced operators. Present day practice moves towards auto tuning, to increase operator comfort. A practical issue is that during tuning, material is wasted. Hence need arises to shorten the time needed for tuning, and the trial-and-error involved, so that wastage is minimised.
In coming months, we will take up a few control challenges, and examine the schemes used for solving them, and some tricks and tips which industry uses.
Sivaram PV, Non-Executive Chairman, B&R Automation India, has been associated with the company since its inception in India. Leader of B&R Education Network, he is currently engaged with Educational Institutions to add elements of Automation into Engineering curriculum of all branches.