End, begining No. 11
Even devices need eyes and ears
Let’ s say a digital camera or a Smartphone that are filled with various embedded systems remain devices with untapped potential or a kind of “black box” in the hands of a person who is not aware of certain features?
M. Viliūnas. Your camera has certain functions that help to make good quality photos even without asking your permission.
And whether these additional functions are convenient and intuitively perceived depend on the level of designers. In the devices of the mentioned type, one of the most important and sometimes essential components is user interface. The money saved when designing user interface will rebound on partly used device functions. Therefore, often when little effort is spared on designing good user interface, a significant part of control is instantly assigned to automation.
Currently, embedded systems cover almost all our life. In many places they have already became a norm, however, there are still areas where their usage is not that common. Sometimes, it is even hard to tell does the device have an embedded system. Let’s take drill as an example. It has a function of adjusting the power, which is accomplished with a help of a controller. If the torque, speed is also measured, and operation mode is adjusted accordingly, this may be thanks to certain embedded system. As a result, more energy is saved, the drill has longer life, and in this way we acquire various advantages working with the tool.
What is needed for embedded system to work reliably? Most probably, it would be workflow feedback, which means that we need a sensor, or even more than one?
M. Viliūnas. We need “eyes” and “ears”: without seeing and hearing the environment, we will not be able to perform the control. Probably the only exception would be timer: setting a time interval during which the device will be operating. Timer most often is also an embedded system, however it is on the lowest step of embedded systems. If you want more functions, sensors are required. They ensure getting feedback. Selection of sensors is an important stage of system design. However, the essential component is control algorithm. In simple systems, standard control algorithms are applied, and the most complex systems are installed algorithms imitating even certain functions of artificial intelligence, for example, when a car can “itself” make certain decisions and drive without driver’s assistance.
Speed is everything
You have mentioned certain shortcomings of embedded systems developed on the basis of amorphous electronics: significantly lower speed and energy efficiency.
M. Viliūnas. There are no processors made of amorphous materials yet, therefore it is early to talk about amorphous embedded systems. So far we only have amorphous electronic components: transistors, diodes, photodiodes, and resistors. Using this type of components reduces the speed of electronic devices being developed. If clock speed of crystalline semiconductors is measured by 1 GHz, then after applying amorphous semiconductors one will be lucky to achieve the clock speed of 1 kHz, i.e. million times smaller. What does kilohertz mean? It means that a thousand of actions can be accomplished within one second. Given well-arranged algorithm, this could be enough for certain devices.
But perhaps this shortage of amorphous semiconductor bandwidth could be turned into advantage in some way similarly as in the case of using them under conditions of radiation?
M. Viliūnas. I’m afraid it is not possible. Speed nowadays is everything. Much depends on design. There are some niches where these slow components can be successfully used. And taking into account the advantages of chaotic semiconductors, the result can be really good.
You cannot develop a good material without accurate measurements
What tasks do you usually perform?
M. Viliūnas. Talking about the activity related to embedded systems, I usually engage in designing the devices. These are mostly process control devices. As one of the most successful devices, I could mention auto-focusing controller which is intended to ensure focusing of laser beam even when the surface is uneven. We have been selling these devices for more than five years for one Japan company that is installing them into industrial equipment for LED production. By the way, we’re pleased by the fact that this equipment also contain Lithuanian laser. Currently we work in the area of designing controllers that would operate in an optimum power mode. They are intended for connecting solar cell modules of nonstandard and uneven voltage to a single route.
Moreover, I work with students. During the courses delivered by me, students are introduced to processors and their application – embedded systems. One of the courses delivered is concerned with problems of measuring.
Another area of my activity is research of materials. If we want to produce a good material, we need to learn to measure its parameters. This is the strength of our department – we know how to measure parameters well. Having the knowledge of how to measure parameters, we attempt developing new materials and devices. Research often requires non-standard equipment, and often we need to design it.
One of the most significant achievements in the Department of solid-state electronics is CELIV (Charge Extraction by Linearly Increasing Voltage) methodology designed based on the idea of professor, academician Gytis Juška and under his supervision. This method together with TOF (Time Of Flight) method are the main ones for researching charge carrier transport characteristics in semiconductors. However, CELIV can be used for researching materials with higher conductibility than it is possible with TOF. Moreover, CELIV has certain advantages in highly chaotic materials, such as especially popular organic semiconductors. Thus, first introduced in 2000, CELIV became popular very quickly and currently is the first option when researching charge carrier transport in organic materials. It is nice to know I have contributed to the development of this methodology by designing models of calculating time-dependence of current in the materials with various parameters.
Talking about the achievements of our department, we also need to mention another invention by G.Juška and his colleagues. The professor discovered interband carrier impact ionization phenomenon in amorphous selenium, which has allowed obtaining thousands of times greater signal under conditions of low photogeneration. This phenomenon was applied when designing extremely sensitive video camera. They can be used for filming both during night and under water. It also may be applied in medicine when there is a need to take pictures of fine blood vessels of up to 50 micrometers to determine the place of cancer formations and the like. The very camera is made in Japan applying the discovered effect in photosensitive level of vidicon.
Fundamental basics are always required
What is the role of embedded systems here?
M. Viliūnas. In discovering these phenomena or effects, embedded systems do not appear, however they are actually used. They appear in the point of turning towards development of a final product. For example, we have organic LEDs and we need to show some image, i.e. to make a screen of LEDs. How to do this?
This may be accomplished applying engineering solutions from other areas.
M. Viliūnas. Yes, because this is the stage where engineering begins. New materials and components made of them is physics, and embedded systems is already engineering.
They support one another.
M. Viliūnas. Knowledge in physics together with engineering skills can give very good results.
However, the background is physics?
M. Viliūnas. My answer would be that physics is the background of everything. I would not take out embedded systems from the general context since this is engineering that facilitates life and improves the quality of all the devices. Currently in physics (and other areas) lasers are very widely used. However, without embedded systems even the best lasers cannot do much. Many parameters of lasers must be controlled, maintained, and usage of laser beam also requires usage of embedded systems. They help focusing, directing, and simulating beam.
Physical knowledge is also required. If a person is knowledgeable in engineering, however has a poor understanding of physical processes to be regulated or controlled, he is likely to choose not the most optimum technical solution. This is why knowledge in contiguous areas is required both for engineer and researchers.
I could refer to the experience of my former students. I do not follow what universities they enter or where do my former students work, but I am aware that from the students whose final papers I supervised, two started master studies abroad, and one started working immediately. All of them were employed is companies working with embedded systems. Interesting selection criteria, I should say. Since our boys have completed physical sciences, they did not much practical experience of working with embedded systems. At least in this area, they find hard times competing with the graduates of engineering high schools. Their tests results were lower than that of the representatives of engineering sciences, however, during live conversations they showed their knowledge in fundamentals sciences. One case has surprised me greatly when one of our students got the job abroad after firing local employee with high education in engineering. Thus, knowledge in physical processes is very useful.
Interest in work being carried out is also very important. If a physicist is also interested in embedded systems, it is apparent he is a motivated employee in the engineering activity of this profile. Even without knowing certain things and not having respective skills, this kind of worker will acquire them very quickly while working.
Interwieved Gediminas Zemlickas