Pathway Seminar Control Systems for Spring 2021
![Pathway Seminar Control Systems for Spring 2021 Pathway Seminar Control Systems for Spring 2021](/pic/pathway_seminar_control_systems_for_spring_2021/eWxwU0VDR0FKSVk_.jpeg)
okay folks uh i'd like to welcome you i hope all of you are are doing well i know that we're going through some very very very difficult times that have uh touched so many of us uh try to get a try to get uh get your vaccine as soon as possible please okay is that required uh is that required by the school now i hope so well anyway today i'm going to be talking about one of my favorite topics uh that would be controls there are um uh several areas uh in in electrical uh engineering one of them is controls uh i must tell you that when i was a student i was interested in many of the other areas i was interested in circuits i was interested in signal processing communications etc but the reason the fundamental reason that i fell in love with controls is because i saw that controls could be applied to virtually every discipline i saw that certainly controls could be applied within electrical engineering in the context of power systems uh i saw that it could be applied within the context of communication systems also within electrical engineering i saw controls concepts within signal processing and electromagnetism etc and so forth but i also saw that it could be applied to for example uh robots and missiles and airplanes and spacecraft and so that really really excited me and so i was hooked uh and so uh you can see why i put on my slide that uh i really love my job and uh for oh about 15 years i've been saying that the age of intelligent systems has arrived and uh that was that was before uh you guys were seeing these driverless vehicles around in these google uh cars and and uh and the quadcopters all over the place and the um the uh rockets that uh that uh that uh automatically land etc i was expecting anticipating all that stuff it came a little later than i thought but uh it is here and it is here to stay folks and there are many opportunities for you and that is part of what i wanna uh talk to you about for the first time in history in history uh there are amazing technologies that are accessible uh to the masses and by that i mean uh uh computing technologies uh uh sensing technologies actuating technologies and uh we can really now build systems that are very very very complex excuse me folks um of course people are trying to reach me now and uh and and despite the fact that i've tried to let everyone know that i'd be busy uh i apologize okay so there's a lot of cool stuff that we all have access to that we did not have access to just not that many years ago to give you an example of what i mean uh about just 15 years ago just 15 years ago we still really needed a super computer and and i for example the nasa columbia super computer that's a super computer that i had access to for designing hypersonic vehicles these are vehicles that go uh at least five times the speed of sound and and so uh you needed a supercomputer in order to design these vehicles but today we can design we can do the majority of the design of these vehicles on our little desktop computers and now even our laptop computers so that's sure that is a evolutionary uh you know it it happened little by little but uh it it but it's sort of in my world that i live in this controls community interacting with the nasa people and the military people in in in industry it really sort of happened just like that and and it was uh not just evolutionary but it was revolutionary because getting time on a super computer that's not easy to do because that's a national resource and and you've got people competing for that all over the country okay so i could go on and on uh making a list of of uh of things that we now have access to that we did not have access to five years ago ten years ago and the world uh has changed uh here to to to tell you just a little something uh just about three years ago i was talking to my students about these new vaccines that were coming uh and and of course folks there's just no way in hell that i knew that they would be coming right now uh and there was just no way that i knew that we would have this pandemic of course and uh but uh like i like i said there are just amazing amazing things that are available now that that didn't exist before and i need you to be aware of these excuse me i need you to be aware of these things uh because that means opportunities for you and something that i will uh ask you to do after you leave this session is to really think hard about what you like what you don't like what you would like to do in the future uh and start working towards that and of course you know i'm here to sell controls uh sure but first and foremost what i like to do is i like to try to help students figure out what they want to do and if that turns out to be controls well then i can hang out with you for a few years or maybe even a couple more decades but if it turns out to be something else well we will help you to try to do the best you can in that other area okay even if you do choose one of the other areas that's okay they're okay areas too all right let's keep it going here i want to thank all my students and everybody that has helped me um okay so uh here's an outline what is controls where is controls used what are the courses what are the controls faculty what are some of the job opportunities and i've already started talking about the ongoing technological revolution there really is a revolution taking place all around us i mean i mean sir i can point out the obvious to you you know the driverless vehicles the drones uh these uh the new vaccines that are that are that we see and let me just state what i believe uh well let me let me say okay we have an artificial intelligence revolution that is that has captured the imaginations of the entire planet i mean uh the top uh 25 companies in the world are all attacking uh artificial intelligence in in a in a very aggressive manner to incorporate it into into their um into their system because they know that um it's not just a fad it's here to stay but but i must say it it as a controls person it really disturbs me immensely when i hear even highly educated people saying that the computers and the robots are going to take controls professor to say that it will take a very long time to replace human beings we have this amazing thing called a brain and uh folks if we use our brains properly uh the computer really is no match for our brain now don't get me wrong the computer can do lots of stuff that we with our great brains cannot do okay but there is just so much that we can do uh that a computer cannot do and uh i think that um it might be oh some 50 years before we really really see a uh a transformer like uh incremental jump you know i i don't want people going away saying rodriguez says there's going to be transformers in 50 years i did not say that i did not say that but uh it will it will take a very very long time for there to to be um a serious uh i i don't think there will ever be there will never be a takeover because there will always be stuff for human beings to do and again this subject is very dear to me because i often like to make a little joke that uh there's artificial intelligence but in controls where we control robots we control airplanes uh we control economies we control whatever and we expect to get very precise predictable and reliable performance well um we don't rely on artificial intelligence we rely on real intelligence and by that i mean the laws of nature the equations that govern the laws etc the laws of physics the laws of chemistry that is what we do that is what we use to do controls okay let's keep going what is control so i i typically use this block diagram in order to describe what controls is about the idea is that there is this physical system other we control people call a plant it's a physical system or it can be an organization it could be anything some sort of a process anything with inputs and outputs if it has inputs and outputs and i'm able to measure stuff then i can control it okay so it doesn't have to be a robot it can be an economy and yes folks it can be a pandemic all right as soon as this uh so i became aware of the pandemic i believe in december of 2019 that's when it started it just started and i was paying attention to this type of stuff but of course i had no idea that it would be what it is uh no one no one did uh but as a controls person i'm very interested in pandemics of course because they uh can certainly get out of hand and they exhibit uh exponential growth and so you need to measure what's going on in order to be able to control it and i want to make a forceful statement about that because as a nation we have not done uh a very good job in measuring uh what is going on in this pandemic and uh we we're still not doing a very good job in in measuring and uh as we move forward i think um uh we we have gotten better and uh we do have the vaccines now and and we are measuring and um we are scaling up and so that's very good so let me let me talk about this feedback loop again we have this system that with inputs and outputs that we're trying to control and we're trying to control it in the presence of uncertainty what do i mean by that well there's always disturbances present that's what that d i and d naught represent and if you're going to measure the output of a system y and you're going to feed back that measurement uh your sensors are always going to have noise associated with them that is just a practical reality you're never going to have perfect measurements so uh so your measurements are uncertain and so here in this diagram r on the left hand side is is what we call and controls a reference command and it represents the desired output it is what we want y the actual output to be we want y to follow r in the pre even though there are these disturbances that are interfering with that following and even though there's sensor noise interfering interfering with that following and so the whole point of controls is to design that controller k that you see in the diagram based on some information about the system you're trying to control p and that information typically is a mathematical model for p and it could be a very very complicated mathematical model for p all right and uh and there are parameters inside that model that can be uncertain you know the mass of your rocket for example the mass of the rocket is changing with time because fuel is being expended for example okay and then again there are these disturbances present and noise present and you have to have some information about the disturbances and noises present uh you can't just be completely uh ignorant about disturbances for example we all know that you can't fly an airplane through a um um let's see i gotta be careful what i say um it it is dangerous to fly an airplane through a tornado okay now it's different if you're in the eye in the middle where and and it's spinning around you and then it's not uh destroying your airplane uh but typically what the airplanes do is they fly above it and then get into the eye and they avoid all that circular destructive force okay so there's some basic stuff that we know uh when when a plane is landing at an airport if you look to both sides of the airplane there are these little things spinning around what are those little things those little things are anemometers uh they are wind speed sensors and there's a lot of wind speed sensors and what they're trying to detect is wind shear because wind shear can be catastrophic uh to an airplane that's landing because you can just lose all your lift so again my point is that you need to have some information about for example the size of the disturbances and how rapidly they mut they can fluctuate okay because you just can't tolerate arbitrarily sized disturbances and arbitrarily fast fluctuations so this is what controls is fundamentally about designing a controller on the basis of uh information about the system and about this uncertain environment and the the the simplest example that i can give is uh for example the cruise control on a car where you're trying to um control the speed of the car and you set the speed to say 50 miles an hour and the control system automatically maintains that speed whether you're going uphill or downhill or whether there's a wind blowing a rear word or there's a wind blowing forward and uh and all of this is model based okay that's why i refer to it as real intelligence uh not uh and i joke around with artificial i joke around artificial intelligence and uh and the joke that i make there it really is a somewhat substantive joke because uh let's see how to say this nicely uh let's say um misinformation spread so easily especially technical misinformation and uh how if you tell people that robots are going to replace them it will scare them immensely and i know that because i had a father uh who uh was a working man he was a window cleaner and he used to always i think he used to tell me every day computers are going to replace us all and robots and uh and he used to say this and he had a second great education but uh but you know he was nervous about it and he he said son study computers we have a we have a relevant question right on point um there's a student named colton he's asking what your personal definition of intelligence systems would be so so uh the the the um so here here's the problem uh that i have with artificial intelligence and it is that uh sometimes the uh the uh let's see the um what it can do is severely exaggerated uh there is a there's a so i haven't been to the mall in a long long time but uh let's see there's a place where they sell cool stuff is it called the sharper image folks i need a young person out there who does go to the malls is it brookstone brookstone right there you go like like brookstone sharper image there's a a a handful of these stores that sell cool stuff and they say for example they advertise this has a neural network in it and i'll be talking about neural networks and this has some artificial intelligence in it and then when i actually and someone like myself who is a controls person and has been doing controls uh for about almost 35 40 years something like that all right um that's quite a bit of time yeah it is 40 years uh yeah okay i think i just want to go to sleep now uh so so with my background and i know how things work and then i i can see that some of it is just um advertising bs and i don't mean bachelor of science folks uh and so that's very very annoying and and why is it annoying here's why controls uh got us feedback amplifiers it got us all the circuits that we take advantage of today and i'm here i'm showing you my iphone if we did not have controls we wouldn't have these millions of circuits in here and feedback systems that are inside this phone let's see we would never have gotten to the moon we wouldn't have the driverless vehicles oh by the way there wouldn't be artificial intelligence and and the point is that controls has established a solid foundation it is a solid scientific foundation that has existed and like i said we got to the moon because of that solid foundation and there was no artificial intelligence there and and i again i um it was model based and very very rigorous and very scientific and that is what this area is about um so i didn't really answer the question you know what is an intelligent uh system well you know that that is somewhat subjective and whether something is intelligent depends on what you expect from it um uh let me give you an example if i'm um and um these these driverless vehicles they're they're amazing but but bc i don't want to talk about it like it's magic there it is really basic science and grounded and feedback systems and controls okay not artificial intelligence okay they may use artificial intelligence in order to sell certain things and and capture people's imaginations uh because if they start talking about bodhi plots and frequency responses and nyquist plots etc and a lot of the stuff technical nerdy stuff that you learn in your classes then people might get turned off although they do say things like neural networks and fuzzy logic so that's nerdy stuff but uh that's just it's just uh been catchy okay there's one other quick thought that's um this student anthony is taking 350 right now random signal analysis and that's kind of our prerequisite for all the controls stuff yeah he's that it's very statistics based um and he's only done probability analysis type stuff and he wants i think you've answered this already you've shown the complexity of controls but he's wondering if you should expect just more probability and statistics in the controls area okay all right that's a very good question it's a very good question so let me give you the perspective of a few faculty so the main the main faculty in controls uh and i'll be in there on slides here and you have copies of these slides i believe uh the main faculty are dr cicales uh uh uh uh dr costa acosta ciscolis uh dr jenny c and there's uh myself and um uh uh there's another professor that i i'll i'll i'll uh talk about let me see i'll just uh go quickly over there and just uh point out the slide to you guys so you can just quickly see it uh so there's uh dr ly myself dr c dr cicales and and uh dr ly uh he doesn't teach any of the fundamental controls classes but he's involved in brain research and uh and i'll be talking about the research of each of these um each of these professors and let me i got i just short circuited myself because i got an interruption here i apologize um okay so uh if you were to ask dr sokalis or myself or dr c uh but mainly dr sikalis and myself um uh have we used a great deal of probability and statistics in our controls work and i think the answer would be and this is over a period of like 40 years so i think that's a pretty good history there the answer would be no we haven't okay now we use probability and statistical models when they are absolutely needed and many times they are needed okay but in most of our work it hasn't really come up and and i and he we've we've both dealt with lots of systems that have a great deal of uncertainty but uh it might not be represented probabilistically and statistically for example i don't uh the mass uh suppose i want to talk about the mass of a system certainly i can talk about its mean and i can talk about its standard deviation and then i can say would satisfy some gaussian distribution yes and for some applications that might be appropriate but for most applications that i've seen in controls and i have seen thousands thousands of applications uh you might describe the mass as okay uh 200 plus or minus 25 and that's it okay that's the way it comes off of the production line you understand so i hope uh i believe i've answered your question about probability and statistics uh no you will not be inundated with probability and statistics okay all right um so there's lots of examples that i can give of uh of of that require controls and on this slide there there are a bunch of uh examples there's the car airplane missiles one of my favorite you see these two airplanes in the bottom left uh the the smaller vehicle that's the hypersonic vehicle so that's a vehicle that goes at least uh five times the speed of sound and anywhere from five to ten times the speed of sound and uh they have been so i've worked on that since uh about 2004 and they just recently within the last two years entered the global uh news because you have the russians working on hypersonic vehicles and and others working on hypersonic vehicles um so um all right lots of stuff requires feedback let me um let me keep going here okay um this some of these slides are a little nerdy and i'll i'll just um skip them all right but uh but you can ask me questions about anything whatsoever uh during this presentation and after the presentation okay i'll be happy to talk to you i'll i could give you my phone number you can call me and we can chat about anything okay all right so controls really is everywhere and it is inherently multi-disciplinary when i so i'm an electrical engineer all my degrees are an electrical engineer but i have a bunch of friends some of some of them initially thought i was an aerospace engineer because i work on a lot of airplanes others thought i was a mechanical engineer because they see me work on a lot of mechanical systems all right others you know thought i was a chemical engineer because they saw me working on semiconductor processes no i'm an electrical engineer okay but i'm interested in a lot of stuff that is not just electrical engineering stuff not just uh power systems and circuits that use feedback systems those are very very uh exciting to me i love that stuff all right but uh i like a lot of the other stuff okay so on these slides i have a very long list of stuff that needs to be controlled and the reason for presenting it is to open up your eyes and your mind and ideally uh get those juices flowing in your brain and get you dreaming a little bit and planning planning what do you want to do in the future and it starts by thinking right now hey what do i like what am i interested in what do i need to learn okay that's what it starts with all right so uh here at the very top here is capital investment that's portfolio management so that that is investing selling stocks buying stocks selling bonds buying bonds uh and and that type of stuff and i've had students do that after they got their engineering degree all right because a lot of the thinking that goes into controls can be used in lots of different fields you just have to see how it can be used okay and and that's very important so that you can sell yourself properly when you graduate okay that's very very very important because sometimes uh i'll i'll see students they have an opportunity right before them but they don't realize that their background applies and so they may miss the opportunity because they don't see the connection and so uh so i talk to students a great deal about uh how their stuff can be used okay all right but this list is pretty pretty amazing uh and again this is why i fell in love with controls a long time ago folks um now what are the courses all right so for me uh uh and for all of you you know we all take our basic circuits class and that's where we start learning about basic circuits and maybe some laplace transforms and solving differential equations and uh and concepts like transfer functions and uh uh it was there where things started for me and then as soon as i realized that that transfer function can be not just for a circuit but for a robot or a spacecraft or a nuclear reactor or a rocket or you know and i could go on and on or an economy as soon as i realized that i said wow oh yeah i think i'm interested in controls okay and then then we all take ordinary differential equations that's ode and uh we learn about laplace uh transforms and we learn linear algebra and we take signals and systems all right and signals and systems is a very very important class very very important uh i recommend that you take uh both of those classes and that uh before you start taking the controls classes uh you should certainly have the first signals in systems and ideally you should be taking the second signals and systems at the same time as the controls classes the fundamental controls classes that we offer are tripoli 480 and 481 these are our bread and butter controls classes i typically teach 480 okay and uh i i i i should uh i i need to update my slides there's another instructor that we have uh within for a within uh um within controls area and um uh uh i call him that's his his name and um uh he was he was a student just a few years ago but now he's an instructor and he teaches 481 and he works on brain research and so you want to get to to know him he's a very nice guy a very capable individual and uh he's very helpful all right he teach he usually teaches uh 481 sometimes i've taught 481 in the past dr cicales has taught 481 in the past uh dr cicales has also taught 480 in the past but these days usually i'm the one teaching 480 and uh ashfaq is is uh the instructor for 481 now let me tell you something about the difference between these classes okay 481 is intended to be the more applied class that is its purpose okay and so in that sense um if you were only going to take one controls class and never do controls again i would tell you to do 481 because 481 addresses continuous time stuff as well as discrete time stuff that means you're going to have laplace transforms that's for continuous time you're going to have z transforms that's for discrete time and then you're going to be taught on how to design control systems okay a little bit of analysis and design of control systems and then you're going to be taught a little bit how to implement the control system in some sort of digital computing device like a computer like a microcontroller okay uh or or or something else you know like uh an fpga a field programmable gate array now usually they focus on um microcontrollers and and and uh uh and computer computing devices in 481 now 481 has a lot in it it is supposed to be the more applied class but just think about what i just said it has laplace transforms as well as z transforms and if 481 is your first controls class then you're going to be learning the basics of controls as well as these laplace concepts and these z transform concepts so that's a lot of stuff to digest with that said my advice to all of you is this if your intention is to do controls long term then i would advise you strongly take 480 first and then take 481 the students most of my students over the years have done that and after taking 480 they go into 481 they know how to design control systems really really well really well in continuous time and so they know all the laplace concepts really really well and so when they get into 481 you know the first few weeks of the semester where you're learning the basics of what controls is about well they already know the basics and then they s then they start teaching in 481 some laplace transform and bode plots and stuff like that well the students that took 480 they already know that okay so it's only when they get to z transforms the the discrete or the digital stuff okay that's when you got to wake up and say all right this is new it wasn't in 480. okay so uh my strong advice if at all possible take 480 first and then 481 it is easier to do it that way because if you get a solid foundation in 480 then you are much more open to digesting mastering the material in 481 okay all right what other classes are are there so one of the big points that that i want to make folks is please go to graduate school now uh those of you that uh that are thinking hey i just want to get my bachelor's degree and run hey i understand i know this stuff can be tough i know that and i know that when you graduate with a bachelor's degree you could have a very very good job with a high-paying salary i know that i know that and that that high-paying salary pulls a lot of our students into industry but i'm here to tell you uh please please please please think about going to graduate school think about the four plus one uh uh system all right uh think about that four plus one uh mechanism uh it is that has been on fire a lot of students have taken advantage of it the reason i want you to go to graduate school and by the way no one is is paying me to say this or anything like that uh i'm saying the same thing to my my kids and my kids are 15 years old all right um the reason i want you to go to graduate school is not because you're going to make 10 to 15 000 a year more although i believe that's kind of cool that's just an added plus the reason i want you to go to graduate school is because you will be able to work on much cooler problems much cooler problems you will be given you will be afforded much more you'll have much more flexibility uh you'll be able they'll turn you loose on stuff more you're likely to be supervising some of those bachelor's graduates uh oh yes and when the layoffs come and layoffs always come in any economy there are always ups and downs okay when the layoffs come they're more likely to keep the students that have or the the engineers sorry the engineers that have the higher degrees and more training so think about it that way but the main reason i want you to get is because you can work on cooler problems and you can have more job flexibility and that is amazing because if you really love your job it makes a big big big difference folks and um okay let me talk about some of the uh graduate courses that we offer in controls there is linear systems that's 582. that class recently has been taught uh by dr cicales all right i've taught it many times in the past i might teach it in the future but recently it's been dr cicales uh uh that is a very important uh it is uh the first important gradual level class okay uh then there is non-linear systems non-linear systems is 586 that class is typically taught by dr cicales i've taught it in the past i believe dr c has has taught it as well but usually it's dr sakalis that has taught that class transform theory 550 that's a class that many of our students have taken over the years i've taught it dr sarcolysis taught it but usually it's someone in the signal processing area that teaches uh that class and that is a a very good class to take for doing advanced work all of these classes are good classes to take for doing advanced work and participating in this technological revolution that's taking place today and this artificial intelligence revolution that's taking place and there i do mean it in a positive way there are lots of opportunities that are arising right now and that will arise within the next 10 years i mean it the next 10 years are going to be phenomenal folks they will be phenomenal so uh controls is uh is a very important area in in it will determine uh to a large extent how we evolve these intelligent systems okay because at the heart of the intelligent systems is feedback and feedback algorithms and models and concepts like transfer functions instability and body plots and if that stuff is not done right then you don't have basic guarantees and as engineers when we design systems we demand basic guarantees that's what these classes are all about to learn the basics so that you can then think rigorously and precisely uh so that you can design things so that uh they behave predictably and reliably we are not hobbyists we are engineers we are not hobbying around we are engineering okay there's also random signals and where you get uh a big dose of probability theory and statistics and we talked a little bit about that earlier there is robust multi-variable uh control i usually teach that that's uh that's my baby in the basic controls classes 480 and 481 you learn about single input single output systems okay for example in an airplane how do you make it go faster well more fuel to the engine same thing for a car etc so that's a single input influencing the output single output namely speed however we know that an airplane and most systems have multiple inputs and multiple outputs that's what you learn about in multi-variable control 588 a car for example you have a steering wheel input to control the direction that you go in and then you have your accelerator pedal to control the speed and i can say the same thing for an airplane for a missile for a rocket for a spacecraft for a robot etc and so forth for a drone okay all right uh there is optimal control let me uh move these pictures aside here there's optimal control optimal control is being taught by dr jenny c uh this semester and and it is about optimizing performance as engineers and hopefully as people we always want to do better how can we make something better well optimal control 587 is fundamentally about how do you do that how do you design how do you optimize a system okay now in real life there really there really is no such thing as optimal usually in real life there's trade-offs for example uh buying a car you know okay there's we all usually most of us most normal human beings set a limit on how much we're willing to pay for a car okay and okay that's going to determine what we can get all right there's a few people on the planet you know like your bill gates and and and and some of those who you know they can uh they can buy the most expensive car on the planet and they can buy uh a hundred of them all right but that's that's that's an outlier okay folks um all right uh let's keep going neural networks yeah could i be rude and interrupt for a minute please okay well we're getting a little close to the end but i wanted to be sure that a question from colton got answered which i don't think we did address okay do you mind if i ask you this question for him please please please please my people wants colton wants to know how does the electrical controls coursework differ from the mechanical controls coursework um so that that's a that's a good question uh i've interacted uh with a lot of mechanical engineering students and professors over the years i'm very familiar with what they teach how they teach it um let's see how to how to say this um well the it it fundamentally depends on the professor that you have uh and what they cover uh you've heard what comes out of my mouth uh i like to address a whole bunch of different systems and and uh aerospace and mechanical etc uh in in the mechanical engineering controls classes i think you will see that they address a lot of uh mechanical systems like uh uh ground vehicles um ground robots uh now we do uh similar stuff uh in art classes so uh if if i if i had to say from an apply i might say that their classes might be a little bit more applied sometimes okay in our classes uh tend to be maybe a little bit more theoretical uh but that that's good and bad on both sides more applied okay you can get more hands-on more theoretical okay you can do more advanced work all right uh but but uh but but we have students designing robots and building robots uh i've i've supervised a bunch of senior design projects where the students have built quadcopters and built robots and won the senior design competition so uh i think that answers the question i think okay any other questions there was a question from seth he wanted to know is the controls focused masters program ever going to be offered online ah that's a good question uh it comes up it's been coming up every semester and i'm here to say that uh that a good answer to that question i don't have the best answer to it because i don't make that final decision that decision that decision is is is made by my superiors all right however uh i think it's coming very soon a lot of our classes uh at the graduate level are already offered online yes i would even venture to say that all of them or or and most of them at least are offered online and i would even uh say that i would think that all of them are um are are are online i think dr zakalis offers all his classes online dr c offers all her classes online and me uh i offer all my classes online so um uh it's just a matter of time before the superiors uh make it official okay so i think that answers your question all right seth another question this one is from chase chase wants to know how much more rigorous is the work in 480 481 over 304 and asking for those that may not be doing the best in 304 but are interested in this area such as chase okay all right so let me let me say it like this so i i think our 480 and 481 classes they're popular a lot of students take them and uh and i can speak more for 480 because i teach that uh so often um if i had to say which is the harder class uh again 480 is supposed to be the harder class but 481 has in a way more concepts in it so it's deceptive when i say it's the more applied class uh it is supposed to be the more applied class but like i said it has laplace transforms as well as he transforms whereas 480 only has laplace transforms with that said let me say the following uh 480 is still the more rigorous class maybe the more mathematical class but i'm happy to say that over the past 25 approaching 30 years uh the students in 480 have done really well and and let me even quantify it for you and this is being recorded so uh but it is history and uh i don't make an effort to try to uh fit a certain distribution when i'm grading uh i pretty much grade straight forward i in my exams i make them available to the students i've done that for years and years and uh a third of the class and that's a pretty large percentage does really really really really really really well in in in the class and by that i mean you know an a a minus you know and then we still have a lot of students getting b's and you know there's uh uh uh there's students that do get c's and there are every once in a while there's students to do worse but overwhelmingly the majority of students do really well and they have a very good experience and what i recommend that you do is look at just talk to one another look at the look at evaluations you can see how students have evaluated me and uh you can cert you can certainly ask me stuff and uh i i i uh i like helping students very very much and so if you take a class with me uh i'm gonna beg you to call me i'mma beg you to answ ask me questions because i'm not good at reading minds i don't know how to do that yet i'm working on it but i don't know how to do that yet you can ask my wife okay here's here's a quick question for you if you don't mind another one um how closely do engineers with the masters in the controls pathway work with plcs okay so that that really depends on where you end up working so i've had uh students that work uh on plc's that have worked on plc's when they graduated and i have others that have never seen a plc okay and so uh but that goes for fpgas field programmable gate arrays and that goes for almost anything uh some students will work on neural networks some will never see a neural network some will use some artificial intelligence some might not even see something that is traditionally called artificial intelligence all right so it really fundamentally depends on where you end up working and uh that that question you can follow it up with uh 20 more questions and in the next uh year or so and i'll be happy to uh help you answer them any other questions yes there was a good one down here let me get to it bear with me for a second thank you so much cheryl of course um this question's from chase and chase is saying with there being only 480 and 481 as senior electives what other classes such as signal processing classes would be the next best for students trying to get as much experience through their undergrad classes excellent so uh i'm glad you mentioned signal processing because i encourage highly highly encourage all the students in the controls area to take signal processing because that is a fundamental that's fundamental knowledge uh to anyone working on signals or on systems and if you're in the controls world you're working on both signals and systems so signal processing is there and and there's other classes that uh that i typically would recommend to students uh that i have recommended uh there are classes in mechanical engineering that i sometimes recommend there's uh often a class on robotics that i might recommend there might be a class on driverless vehicles that might recommend and uh i don't put these on my slides because these are new classes and uh because uh i do want students to sort of try to figure out what they are interested in and then they can talk to me and then i can help them out and of course you don't just you can you can talk to any of the other professors folks you can talk to dr sakalis dr c dr ly uh i need to say that dr cicales he works a great deal on semiconductor research dr c works a great deal on brain research and neural networks okay and dr ly i already said he works a great deal on uh brain research oh dr sakalis has also worked on brain research all of these three have worked on brain research and that includes uh even using uh doing experimentation that involves animals okay and so it could be very very uh interesting and forward-looking uh really revolutionary work because again in the next 10 years we're going to see amazing things happen one of the most amazing things that are growing that's going on right now is this uh we are growing organs in the lab we are repairing organs we know how to repair organs and we know how to grow organs this is cool on another level folks this is phenomenal uh it really is and it is to say that it's revolutionary is almost an understatement it is a severe understatement because we're talking about creation of life and uh that is going to impact us all just within the next 10 years any other questions yes there's a very good question here from chase do you or other controls professors do individual studies classes yes and i i that's good i'm glad you uh uh brought that up so i strongly recommend independent uh study classes where you sit down with a professor you essentially make up uh an outline of what you're going to do and you decide how many credits it's going to be and usually all you got to do is write a report that depends on the professor i haven't seen professors give exams in these classes but there i'm sure there might be some but uh i usually do these types of classes with students after i find out what they're interested in and then i try to uh tailor the class to their interests so that the class can help them with their senior design project or with their honors barrett thesis or to prepare for their master's thesis okay uh i didn't mention a master's thesis uh most of the students that uh that do uh stay for a masters they just choose the exam route uh and i believe that that is gonna turn into a portfolio soon and however there is a thesis route where you uh take two less classes but you have to do a thesis and my advice to students always is if you really want to learn about something at the graduate level do a thesis because then you can really uh put a lot into that and then you can have a document at the end that you can present to the employer uh pres prospective employer and say hey look at what i did and that could be very very very cool uh having said that a lot of students i think are afraid to do a thesis and because they haven't done that they've done exams in the past they feel comfortable or more comfortable i should say with exams than doing a thesis any other questions i believe that's it so far i hope i've gotten to everybody's questions who's posted them in the chat room okay uh can i have just a few minutes to just uh finish i believe so okay so then there uh there's uh uh just a couple of other classes that i need to say some things about and uh there is filtering of stochastic processes that's 581 that's typically taught by dr c uh adapt and that's where you'll learn about kalman filtering and so uh in there you're going to get some more probability and stochastic processes and again i would recommend that everyone take that because that is so important uh the common filter played a a significant role in getting us to the moon okay uh then there's adaptive control that's 686. that's dr sakalis's uh uh baby at the graduate level and then there are other classes that our students have taken over the years their system identification that is typically offered by professor dan rivera in chemical engineering yes chemical engineering and our students have taken system identification for many many many years and then in in that class you learn uh given inputs and outputs what is the system in there so you're trying to figure out a model for the system given knowledge of inputs and outputs and that is so fundamental moving forward in the uh technologically advanced world that we live in today uh then there is applied optimization that is typically taught uh by professor middlemen in the math department and then there's numerical analysis also taught in the math department and so again these are classes that traditionally our students have taken over the years and i said uh i said uh before that uh most of our graduate students when they graduate they have taken the mse exam however i believe that that's going to change to a portfolio uh because lots of uh schools and disciplines are moving toward portfolios but i can't emphasize enough what i said earlier namely that if you really want to learn about something the way to do it at the master's level is to do that thesis and yeah some students they do that thesis and then that leads them to think hmm i want to hang out here for some more years and then do a phd and then they do another thesis and we can talk about that at some other time but please get your master's degree folks and that is not that is not the nerdy professor in me talking that is more the down-to-earth uh dude in me i want you to make some more money but more fundamentally i want you to love your job and if you've got that master's degree uh you'll have much more flexibility and you can work on much cooler problems all right i i've already said something about the faculty and what they work on so i'm pretty done with that and folks there are a lot of job opportunities you just have to realize how this stuff connects to those opportunities and that's part of the education that uh that you need to uh dive into all right you need to see where that artificial intelligence stuff is being used and then you can say hey i have a background in neural networks common filters uh feedback systems uh etc and so forth and optimization and then oh and you may have taken a class uh in computer science on artificial intelligence someone asked earlier are there any other classes that you can take there are a lot of other classes that you might be able to take at the undergraduate level and at the graduate level it's just a matter of checking it out and uh figuring out what it is that you like okay there's a lot of classes folks there really is a lot okay uh so i'm pretty much done uh i think controls is a great great great area i know i'm a little biased uh are there any questions any other questions folks we've got one here the question from rollin is is there any courses even the ee program that cover quaterninians if i'm saying that word correctly i think i think that uh it's uh quaternions quaternions and it continues use specifically in orientation avoiding gimbal lock etc good okay so um uh i know uh in in um sometimes in in multi-variable so in multi-variable control i allow students to pick systems that they are interested in learning about uh and they have to learn how to model it and they have to learn how to control it and certainly some students over the years uh uh their project their semester project has involved quaternions all right but quaternions typically is taught uh in a class uh in an aerospace uh uh class uh traditionally or uh in a mechanical engineering class where they're talking about uh uh gyroscopic phenomena uh and then quaternions come up uh etc as well as gimbal lock and and that stuff uh but uh um it probably i think it's fair to me to say that whether you take classes in electrical or mechanical or arrow you've gotta make an effort to really learn about quaternions yourself uh i've had some of my uh graduate students uh learn about quaternions um uh and they uh we were working on spacecraft and satellites and so it was just natural to use quaternions and um and uh but they picked up uh papers and theses and did it themselves uh i wanna uh uh inject this message the purpose of all this education really is not just to accumulate a bunch of facts that's part of it and that will lead to this high-paying job but much more fundamentally the purpose of this education is to learn how to learn when you graduate that's when you're learning really really really begins don't let that scare you that's actually exciting you're gonna learn a lot of stuff and all of this stuff you learn in school provide you with a foundation so you can learn about anything that you want like quaternions okay anything else that looks like that's the end of questions right now okay well then uh i think that uh let's see i went 15 minutes over or a little bit over that yeah love you all if you wanna uh speak to me uh anytime uh uh cheryl i think you have my phone number yes you can give them you can give them my cell phone number okay that is fine with me okay and uh folks out there you can text me and uh as long as my boss doesn't have something for me to do that would be my wife and uh or my other boss uh at asu uh then i'm here to help you out okay you guys you folks take care stay safe stay away from uh people that cough too much and uh get vaccinated thank you so much for doing this presentation sure uh my pleasure cheryl and thank you for always being so supportive you're welcome you take care okay yes all right then uh i will just leave the meeting right yes
2021-02-18 19:31