 What was the last talk of our session, of our track? Wolf of Power is going to talk about the Great Education, which is a bit backstopping, I can say. Thank you for coming. At S3 there's a 5th or 8th show, which actually is a very nice one, if you would see one, I will comment that. Let's give a round of applause for Wolf. Thanks everyone for being here, especially on a Sunday afternoon, on the 3rd day of the conference. That was a small mishap, actually, with this extra arm. For whatever reason, the abstract did not get put up in the program. So if you look up the program, you see my entire biography, with that platform, there's lots of interesting parts. Now the thing is, I was actually hoping to get some teachers in here. Are there any teachers active? But you're probably an enlightened teacher already, right? So I'm probably preaching to the choir here. So I'm probably going to tell you things that you already all know, so it's not really new. You can also guess it's not really new, because if you look at my hair, I can estimate I'm already a bit older than the average participant here of the conference. So my views may be a bit old-fashioned, but, well, let's see. Okay, I'm with the Science Center. Science Center is a government organization. My boss may not necessarily agree with what I say. And I'm sure our IT people definitely don't agree with what I say. But yeah, no. In Science Center, I work in a research lab called Cradle. And if you're familiar with the local education system, it's very much focused on road learning, textbook learning. What we do is we provide science experiments on experimentation. We do research projects with students to give them an opportunity to see what all these concepts actually mean. So it's not just for the exam, but actually for the privings. In my experience, I was a big gap. Basically, we have students who can recite all the formulas, give them a simple problem that cries the application of these formulas and they don't see the connection. That's basically what we try to, the gap that we're trying to fill. In the process, then we design a lot of our own experiments that we do in our workshops. We do a lot of prototypes in our research projects with the students. And it's an excellent way actually to understand the meaning because Richard Feynman once said, what I think was found in this death room on a blackboard, what I cannot reproduce, I do not understand. The best way to learn something is to build something that uses it because then you know or you can make sure and you really understand it or you see whether they have misconceptions. For the same reason, then we also like to use our home design, home-built equipment in science experiments. We don't want to do by the usual commercial set simply because if the home-built we have complete control, we know how everything works. And the students and teachers alike can see that things can be home-built and can actually work. So basically it's about removing the mental block that things are beyond the ability of normal people. And of course we use computers for our R&D, which is what this talk is about. Okay, and some personal observations. I'm possibly the oldest person here in the room, quite likely. So I went to... Okay, I have to stay in the picture, is it? No, no. Never mind. For my belly, you need a wide angle anyway. So I was in high school in the 1980s. These were the old days of home computers. So we had 8-bit computers. One of my first computers was a ZX81. Anyone still remembers that one? 1 kilobyte RAM. Was not enough to fill the screen. But it came with a manual that basically taught you things. In those days we had mainstream magazines that discussed algorithms, computer science, electronics behind it, and so on and so on, programming languages. In math class, I'm not quite sure, but it has to be one of those because I remember the teacher. We learned Euclid's algorithm. We learned Christina's prime-damper sieve. Classic algorithms. The same math teacher brought in his computer to show us how all this translates actually in doing math. He had written a program to demonstrate how to bear with medical fractions. So it kind of makes sense. My physics textbook actually had basic program code to do scientific simulations in there and explain how it worked. So these were the 80s. In our days, I mentor students here in our teaching and research lab, and they come in and they come in with a fancy book or whatever, so 64-bit notebooks with a few gigabytes of RAM, and all they can do is a bit of what? Not text processing, but what? PowerPoint. And they're really good at PowerPoint. They don't have these boring text lines that I had. They can enter numbers into Excel. Sometimes they can enter formulas and do an ugly graph, but okay, and they're very good at consuming. And when sometimes schools and their students do attachments for Excel, they don't expect any programming or computer skills or whatever from them. So I wonder why. And actually, if you look, how have actually computers changed? Again, back then there were very limited resources. And the limited resources in the software was limited to everything. It was very difficult to do serious stuff, but actually people aspired to do that. People in the 1980s started to look at how can you program computers to simulate things? How can we use computers to design circuit boards? Whatever, and it was a fight because computers didn't have enough memory for a lot of these things. And today it's the opposite situation. We have workstation class computers, even our handphones are more powerful than supercomputers of 30 years ago. And yeah, what do we do with it? So there's a big mismatch. So why can't we empower our students to make better use of computers? And the point that I want to make in my presentation, and I think this is the point that you already all know, it's actually not an issue of the software, or the cost of the software, or the quality of the software. And what I want to show are just a few examples how we use free software in our lab. Okay, we only have enlightened teachers here, but one thing that I assume, right? Are you enlightened? Of course, otherwise you wouldn't be here, right? But a case in point, actually sometimes how we teach or how we talk about things determines the attitudes. So I have a car driving license. It's not a Toyota license or Mitsubishi license. It's a car driving license. But students, they do not learn work processing or presentations. They learn world or PowerPoint. So if this is how we teach them or talk about things, we basically take away the idea or the imagination that there could be something else. And if you think about 1984, the government designed a language to make it impossible to think negative or undesirable thoughts. This is pretty much the same thing. So the one thing is that don't just treat the thinking of your students. By using a market-dominating brand name. But the other thing is also don't just treat yourself about looking or thinking about alternatives. Okay. Now, most of you will be familiar. The academic environment. This is actually where a lot of things come from. Why? Academic environment. It is used to be about creating knowledge understanding. So there is no requirement for commercial viability. It's not entirely true anymore. Nowadays, there's a big pressure on commercializing things, especially in Singapore, I guess, but also worldwide. So it's a bit of eroding, but academia has a tradition of not just creating but also sharing them. So this is why it's such a such a good treasure chest of innovation. Sometimes research is publicly funded. The implication that the results also have to be made available to the public. Example is not exactly academic, but a lot of US government work is in the public domain with a rightful argument. It's paid by the taxpayer. So it belongs to the taxpayer. This is one aspect where the United States are remarkably enlightened. Okay, but anyway, this environment so free software thrives and it trickles down from these environments. That's why we have Linux nowadays. That's why we have a lot of high quality software, but of course the drawback is there's no commercial marketing behind it and academia has little interest in polishing something to make it look glitzy. So which means the owners on finding a software is then not in some marketing agency, but it's on the user. So this is maybe one of the reasons why we are not making or by the open source of free software community is not making as much inroads in the education segment as we might want to. But okay, for for those unenlightened teachers basically just keep looking there are solutions for almost everything and quite often they actually better. Okay, and this I just going to show some random examples of software that we use in our lab and actually some of it is very ancient software and despite the fact that it's ancient no one, the teachers or students do not have do not seem to have heard of it. Okay, who does not know Nuplot? Everyone knows, right? Okay, just for the benefit and let's see, now to find my shell. Here it is, yeah. Okay, in our lab students take data they take data at school, they have all the school resource projects, so let's see. And then they try to plot this in Excel and try some strange spline curve fitting or whatever and call this data analysis. Okay, this is supposed to be a damp oscillation, it's of course very noisy. But how do we just extract the parameters? Has anyone ever tried to do serious curve fitting in Excel? It's almost impossible. Well, you have to be a methodist to try it, right? On the other hand, I'm probably not telling you anything new, we can just define our model function amplitude times an exponential dk, I need my glasses for this, times an oscillation with a frequency and a phase shift. Let's start with some random value. Okay, of course it doesn't match, but it's trivial to do a fitting new plot. And by the way, this is not a high-end computer, it's a cheap netbook, was the cheapest computer I could find about three years ago. And it's a pretty decent fit. And I show this to students and the draw drops. They were not aware that they can do this and it's such ancient software. So I suspect most teachers also don't know this. And just okay, it's great for data analysis, but it's great for documenting, but it's also a great mathematics teaching tool. Of course, there are educational plotting programs for functions for mathematical relations, but actually nothing is as quick and easy as this. You can explore functions, we even have our coordinate system, you can put rulers in there, so you can do measurements, everything. So, yeah, it's one of the very often used tools in our labs, whenever we deal with data. Okay, but you know this, of course. And I have to get back to my talk. Why doesn't this work now? Okay. Why doesn't my talk work now? Okay. Don't tell me that Libro, Libro office things. I don't know here. Okay. So, Maxima, you all know Maxima? I'm giving this talk. Okay. Who does not know Maxima? Okay. No. I keep it very short because I'm not very good at it. Interesting things. Maxima dates back to the 1960s. But most students will not have heard it. So, Symbolic Mathematics, Symbolic Algebra system. Okay. And let's say your students long calculus. Well, let's integrate a simple function. Okay. Easy function to integrate, right? At least it should be. And it's tedious if you do it by hand. At the moment, I show it that doesn't work. So, I think if function I wanted to integrate was actually one here. There we go. Can be done by hand? I'm pretty sure it's accurate. So, the point is some things are within the reach of students to do by hand, but they extremely tedious. With the right software, you empower them to approach not difficult, but tedious problems. Taking the grand work away from them so they can focus actually on the quantum interpretation of the result. What does it mean, actually? Okay. Let's do some physics. Okay. Location is a function of time, right? Does anyone still remember what is velocity? Is the derivative of location with respect to time? What is acceleration? It's a derivative of velocity with respect to time. Oops. And why doesn't this... Sorry, this should not be an equal, this should be a... Okay. Now, we know let's say we do a harmonic oscillator. We all know Newton's law, right? Force equals mass times acceleration. And in the case of harmonic oscillator of a spring mass pendulum, what is the force? It's minus the spring constant times the location. There we go. And of course it's a simple differential equation to solve, but it may not be a simple differential equation to solve for students. Because typically differential equations you learn maybe at A-levels, you learn at university. But no problem. They can see actually how the solution arises. Once they learn to trust. Okay. We need to make a distinction depending on the parameters. Well, we have to make sure our spring constant is positive because otherwise we won't get oscillation positive. And there's always all oscillations either sine component, cosine component with a frequency square root k over m. That's what they learn from a physics textbook but they have a way now actually of verifying this themselves. Okay, I will not dwell too much on this because frankly I'm not very good at maxima for a reason-reason I switched to something else which I hope I'm going to show you as well. Okay, Libre Cat, everyone knows? No? Who does not know? Okay, at least 2 or 3. Okay, it's a simple it's not commercial quality it's a simple 2D cat drafting program but frankly it's good enough for most of what we do in the lab. Let's see where's my DXF In the lab we use a laser cutter to manufacture most of our experiments we use Libre Cat to design it. So here's the design for an experiment for measuring the speed of light it's a it's a multi-part laser cut thing it's actually it's an optional precision mount for mirrors, laser bio it's photo detectors, whatever no problem designing all this and it's free. So you give students the possibility to design things, draft things they get to apply the knowledge they can actually build experiment if they have access to let's say a maker space laser cutter, whatever CNC machine they can design things but it's not just about really designing things I find that also it's a great math learning tool let's create something for example, how do we how do we create, how do we construct a tangent to a circle a tangent starting from a point has anyone ever thought about this it's a classic construction problem how we would do this with a pencil and compass let's see we take a connection from our point to the circle we construct another circle around it and our tangent so it's actually a very nice way to do a constructive geometry what we typically do with a pencil a ruler, a straight edge compass except now we do it with high precision on the computer once we do it with high precision on the computer it can do much more complicated constructions and to start to actually realize how these methods work okay I'm going to run already behind time so let me show some other things electronic circuits you're all familiar with GNU EDA GNU electronic design automation everyone's not in here let me show a very quick one here what we can do we can draw our own circuit diagrams and this is a complexity of the circuit level diagram it's typically what kids at least my day built in primary school the problem was they do not necessarily understand how it works but that's not a problem because it's actually quite difficult to understand how this circuit works because at the moment we have semiconductors involved the equations become very non-linear but we can simulate them so we have simulation software okay let's do a transient simulation of this circuit here's our output and we see this circuit will generate an oscillation it will turn something on and off so this is the voltage at the collector of our circuit so it will turn an LED on and off so we have a corresponding base voltage basically all the equations governing the circuits are solved although they are non-linear equations and this pretty much describes what our circuit will do so students can explore this without having to go through the detailed math and electronics is a topic that's coming back in Singapore in the curriculum now of course if you really want to you can also do your own circuit board here are our components let's just quickly this is routed you can print it out edge it if you want or send it to a fab service so it's enabling technology for kids again to take their theory they learn from the textbook to actually do something useful that actually works and again we also use this ourselves to design our experiments okay now the ultimate thing of course is programming this is still that I think every student should have it's become easier than ever especially my absolute favorite these days is Python I think it's pretty much a favorite of every one often a few lines of code will do the trick and another favorite of mine is actually Jupiter which I only figured out a few weeks ago it's it's a replacement for iPython it's kind of a mathematical like a front end they can run Python code and let's see and it runs in a web browser the interesting thing is you can intermix code and commands you can use a mathematical location for the commands you can write formulas whatever and this is the reason why I use Maximal less and less is actually a very nice computer algebra system for Python called Simpy so again we can solve the same problem we define mass we define the spring constant yeah for velocity we define our acceleration Newton's law and all this is kind of a textbook like format actually that students could easily go through and explore themselves and change things let's construct the equation of motion there it is and here I added actually a damping term some viscous friction and again let's solve the differential equation there it is and let's calculate the result numerically and display what it will be and indeed we get a dent oscillation for the system so this is a spring mass system that has viscous friction in there so you see it corresponds to the textbook theory but it's actually if you look at it it's just a few lines of code it's an interactive session where students can actually use the computer to derive things in a systematic and self-documenting way and the can I overrun? because I was probably 45 minutes originally ok the interesting or one of the things where actually students struggle in sciences is the number of problems that can be solved using simple mathematics that can be solved analytically is actually very small, most problems do not have analytical solutions where the consequence they never show up in textbooks and this breeds the perception that basically science or textbook science is not applicable to the real world and in this case we can do numerical simulation which is way too tedious to do by hand but thanks to computers nowadays we can do it and the simplest example of something is almost impossible to solve by hand it's just a simple pendulum, I swing my arm here this is a non-linear system you can try solving the differential equations you will not get very happy but we can solve it numerically using numeric integration and all the packages are available in Python, it's part of the sci-pi package so here we define just in three lines the entire physics of the system from then on we just specify the parameters how long is our pendulum, what's the acceleration of actually I'm simulating our two pendulums simultaneously so that you see how this may differ from the simple harmonic oscillator that students learn in school let's plot the result, hopefully it works and we get a result that is not what students learn actually the frequency of a pendulum depends on its amplitude we have a larger amplitude, it goes slower and actually immediate application if you have a grandfather clock if you wind it up if you wind the spring up it tends to oscillate with a larger amplitude actually then it goes a bit slower as it runs down it speeds up a bit but not to be discussed in the textbook because it's too complicated of course another thing that you will not get with the approximations that we do in the textbook what happens if you give your pendulum a really big kick your motion is no longer periodic it's clear, eventually you start rotating around at an increased angle you go round and round and round, you never oscillate back again you will never get this from the simplified equations you find in the physics textbook simulating this numerically is trivial if you have a computational power and we have it nowadays okay now this is not entirely programming so let me show two examples about real programming and this was done by a student in the management program that student programmed a complete rate racing almost complete rate racing package for optics from scratch in python it's not the fastest but it's based purely on the science in school textbooks law of refraction and analytical geometry you calculate the intersection of an optical ray with a surface for example then you apply Snell's law to find the refraction this was a two week internship project we had a four week internship project following this one to design an optical spectrometer that made use of this code and within the four weeks basically from zero knowledge we were able to just take the glass and component specifications from optical manufacturers to complete simulations design our spectroscope accordingly and it ran on the first try with a expected or predicted performance and let me just show you what this looks like okay this here is just a demonstration of the rate tracing for example students learn in school that an optical lens focuses parallel light into a point so here's our lens you see the intersection to the lens this is the focus point it's not true there are aberrations a plain lens does not foreslide in one point that's why if you buy a camera lens it has made from so many individual lenses to correct or our image is not created or projected in the plane it's actually on a curved field a curvature of field again all this is done by secondary school students chromatic aberration you can get chromophringes if you have a cheap lens you can get it again here the magnified version you see that the red focal point the green focal point the blue focal point are in completely different places how do we correct that this is something what it gets into advanced optical design we built an achromatic pair we add a second lens and make the second lens partially correct the flaws of the first lens so two lenses here and look at the focal point the focal distance here is now basically the same so we corrected the chromatic aberration so this is actually but not with this low software but this is a method how actually industrially or in real world optical designs are done and this is all school textbook knowledge just augmented with the power of the computer so it's perfectly within the reach of secondary school students the last thing that will be shown is the simulation of the spectrometer itself we're going to show in a few seconds there it is we have an entrance approach so all the light passes through one point goes through a collimating lens goes through a projection lens and we get a spectrum projected here and of course it's not just a prism that has chromatic aberration and splits the light or deflects the light depending on its wavelength the result is actually that the focal plane the surface on which we get the light focus is actually very tilted with respect to the optical axis which is again something that you will not find in the textbook but it's essential actually to design a spectrometer that actually works so we use this simulation actually to find out where to put our photo detector that which is the spectrum okay and I'm already waiting so let me just show you the resultant spectrometer I noticed someone else at the conference was actually showing a spectrophotometer so it's a quite similar principle okay where are we okay and again this was I think 16 or 17 year old student okay it's fairly dark let's see what happened now okay here you see actually the spectrum of our room light let me make it a bit slower let's average a bit and can actually start to identify the spectrum lines this is mercury mercury in here this one is Europium I have to give it a bit more light here a bit more exposure time the thing is in terms of parts the spectroscope costs maybe 120 dollars or so the resolution is basically what I get for a few hundred to a thousand dollars from commercial providers and it was designed and built by 16, 70 year old student within four weeks with no prior experience so these are things that are possible and all this is completely using free software okay now I may over time and again I'm not telling anyone anything new although I may because I was during the keynotes during the opening of the conference and I was a bit shocked how people started talking oh we're all open source, open source and the implication is so free and actually open source has nothing to do with free and the best example is a patent a patent is by definition open source it's by definition not free right moving on a DVD it's open source you're entirely free to receive the content that doesn't mean you can use the content anyway you like right so open source is not free and free of charge is not free as in freedom and actually yeah free of charge is an obvious short term incentive but it can be dangerous and it's actually a marketing strategy that's what people assume by commercial companies in the educational sector they get you in with free or very low cost educational licenses they make sure you get used to the project generate a lot of data files whatever get locked into the project and then you graduate or you want to take a product commercial and all of a sudden students find oh I made a fork out a few thousand dollars for MATLAB for example so for this reason alone I would make an argument it's actually an ethically very difficult decision to train students with educational license commercial software in school the school as a principal as a teacher you can make that decision but it's not a decision that the students may be aware of or that they can fully appreciate in terms of the consequences so I would strongly argue for using free software so the students can take actually something along with them and keep using it okay that's pretty much it and overtime for that my entire point computers are extremely powerful tools so if they're so powerful tools we should empower the students to make use of them they're not just consumer appliances they're really tools students need to learn tool skills okay there's lots of excellent STEM, science, technology, engineering, mathematics software out there it just makes a polish so so we have to look around we have to invest a bit more into learning into the learning curve but in general I think it's worth it makes it practical to take the learning of the students to the next level not just about computers but about mathematics, physics, chemistry whatever because the power of the computer allows them to take the textbook concepts and apply them to real world problems otherwise too complex to handle notice for teachers, students can do amazing things the main problem or one of the main problems I find in our lab we try to offer advanced science classes to the kids they make them measure the charge of the electron speed of light and the reaction from the teachers oh this is too complicated for our children we cannot do this teachers, if you're here students can do, let them do it don't deprive them of this opportunity free of charge free to you and open source are three separate things and keep in mind the ethical issues if you get kids hooked on proprietary commercial software in school that's pretty much it thanks for your patience have you missed the file tornado already I don't know is it 330? ok thanks for staying so long anyway the state of use of this type of software in school for you teachers in English is it very good not in the mainstream you don't know there are some independent schools semi-dependent schools and they tend to higher people outside from the convention education community so they take academic speech what they were in people with research experience this may thrive a bit but by and large the state schools government schools very homogenous Microsoft dominated teachers actually are anything for teachers here teachers are notoriously overworked but let's put it this way a lot of teachers in secondary school are engineering graduates science graduates whatever but actually they don't really know they never program the computer or they never built anything I'm not quite sure how or why but this is the way things are it's a lot of it's an absolute battle basically to change this mindset few years ago actually Ministry of Education in Singapore started an initiative called ALP Applied Learning Program to force secondary schools to teach the kids some hand-on skills and I was involved in a meeting with one of these schools the principal introduced us yes this is our teacher got a degree in mathematics degree in physics, degree in chemistry degree in mechanical engineering electrical engineering this is our committee we have no idea what to do tell us what to do you have the exports there just use what you have university but this is a reality you're from Australia I guess I am from Australia but I'm from Indonesia how do things compare with Australia or Indonesia maybe both Australia is probably the similar situation I started to see before there's a lot of mathematical solutions there does seem to be a pushback that's come through mainly when it's more of a project there's a lot of technical engineering work with brands there's a little very brand-related environment and that has there's been some appreciation that has caused problems when you have people coming out very skilled but I'm still being one approach that I'm foreminded and yet in Indonesia a better way of explaining it is to actually do ICT program in high schools completely there is none very important dominated or otherwise it doesn't exist so I see students in the streets that's the only thing they can do I mean let them comment on the clean the ICT program it's all something that it sounds to me there are ICT programs but they somehow are supposed to be down to here don't hope to make a web page over there and when I think back when I was in 5th or 6th grade no one talked about ICT but we learned about algorithms as part of the standard mathematics curriculum so I wonder where this is going another question let's wrap up the track and then you can ok thanks all for coming today for the track let's give another round of applause