 Good afternoon everyone. That's a quick outline of my talk. I just want to warn people there are some diagnostic medical images throughout the talk. This is part of the content of the talk. They're all false color. I'm not trying to be gory and freak anyone out. So just to set a little bit of context. Surprise, I'm an Australian. If I am difficult to understand, you're going to have to put up with that. If I use any Australian newsrooms, however, do ask for some context there. I am very lucky that throughout my entire career, I've been able to work on and with open source. And a lot of that has been helping researchers among a lot of government funded bodies. Essentially, a lot of the research foundations that I work for, once they've paid for the hardware and the staff, there's no money left over. So we're using free and open source software. I do tend to do a lot of training at different conferences and events on software carpentry and structure, for example. A very painful hobby is running open source conferences in Australia and helping to run them. I ran the Australian Python conference two years running in my previous home state. I'm now on the overview board of running Python conferences in Australia. I'm also helping to run the Linux conference down the Gold Coast. Next year, our call for papers is currently open. If you like being stuck in a plane for 24 hours and then sweating profusely for a week, have I got the conference for you? So, to also set the context, what am I not? I'm not a physicist, a scientist, a software engineer, a researcher, an academic or an electrician. I'm just a software engineer. So there's going to be a lot of content in this slide show that I understand it in a certain way and the technical details of that are almost certainly incorrect, but I hope that the information will be portrayed accurately enough. And fortunately, at a large research institute like the Synchrotron, there are experts in all of these fields that I'm not. So this is a wonderful diagram which shows the entire light spectrum. The bottom bar, you can see the colors that our eyes are actually able to detect in the visible light spectrum. And if you go further to the right, you eventually get into the X-ray spectrum. So you can see that the X-ray spectrum, the waves are about the same size as atoms and molecules. So you can use the X-rays or light waves at that size to tickle atoms and molecules. And that's essentially what the Synchrotron does. And having a closer look at just the X-ray part of the spectrum, you can see the different use cases that we use today for X-ray. So there's medical, chemistry and security uses showcased there. So a Synchrotron, in ways that I can get across if I was doing a tour, it's a very big research tool to look at very small things. You can think of it like a microscope that uses X-rays instead of visible light and sort of remembering that X-rays and visible light, they're on the same spectrum. So they are kind of the same thing, but they're of different sizes and matter reacts differently to X-rays than they do to light. It's a really fancy X-ray machine. If a very fancy X-ray machine and a very fancy microscope had a child, it might be a Synchrotron. But don't repeat that too loudly. So in essence, the way that we generate the X-rays that we use to look at samples, in the two rings that you see there, the Synchrotron accelerates electrons to essentially the speed of light, just show us the speed of light. Regular particles when you accelerate them, they'll want to go into a straight line. We force them to go into a circular pattern. And every time you force the electrons to turn, they get rather grumpy and they throw off an enormous amount of X-rays in the direction that they wanted to go in originally. The amount of X-ray energy that gets thrown off the Synchrotron is roughly a million times brighter than the sun. All of that X-rays that gets generated, it's a wide spectrum of X-rays, all different sizes, all different strengths and all different numbers. So what happens next is each of the X-ray beam lines that is hanging on the outside of the outermost spring there filters the beam down to just the part of the X-ray spectrum that they are particularly interested in using. Then it gets focused onto the sample that they're looking at and it will affect the sample somehow. How that affects the sample is very different depending on exactly what size and energy of X-rays have been filtered out. So in each one of those hutches around the circumference there, you've essentially got some form of filtering going on. This is a fairly standard setup for the beam for each beam line. However, every beam line is different because they're using different parts of the X-ray spectrum, so they'll filter out different parts of it. Because the X-ray beam is so dangerous, none of this is done by humans, all of this is done remotely. So every one of those slits and mirrors and focusing beams there, they are all controlled remotely. So there's tons and tons of motors, sensors and detectors in that setup there, and that's what we, those motors, we use those to control the beam line and how it hits the sample in the bottom right corner there. So it's a silent movie, so you have to listen to me, unfortunately. So all I did here is I took a sports camera and walked around the outside of the technical floor. So you'll see the red and the white there. There are two concrete foundations in the building. We have the white concrete foundation where the synchrotron is sitting on, and then the building that surrounds it is on that red foundation. So if we get a lot of wind up against the building, it might knock the red foundation about, but it's isolated from the white foundation. Because a lot of the experiments that we're doing are on very small things and we can't get them knocked around. So I think we just went past the Australian Cancer Research Fund beam line. So they've actually stumped up the cash to help build one of the beam lines. So any cancer research scientist around Australia can get time at the beam line for looking at particular samples that they need to. When a researcher brings their sample on site, often they have to do a lot of work on the sample to fix it before they can investigate it with the synchrotron. So there's a lot of chemical stores and there's a lot of physical stores on site to do that work. There's always a lot of experiments happening looking at new ways of using the beam line. So that's why I've got those temporary setups there. They've bought or borrowed a new piece of equipment and they're working out if and how they can use it with the synchrotron. A lot of the sample work gets done on frozen samples. So there's a lot of equipment that gets cooled with liquid nitrogen and a lot of the beam lines you actually have to cool the samples down with liquid nitrogen so that the temperatures are low enough that everything's not jiggling around. There are essentially two different colors that you're seeing. I'll go with puce, I think it's called, and the white. Those two colors symbolize the two different types of material. The white is cement and the puce is lead. So the idea there is if something really bad goes wrong and the x-ray beam does not follow where it's supposed to go, it'll hit the lead or the concrete and that lead or the concrete will be thick enough that it will absorb all of that x-ray and it won't hit a human. There's a lot of safety features on the synchrotron. Practically every door that could expose a human to harm at the x-ray beam line is lined up with a safety sensor system so that if a human is in the wrong place we dump the beam line into lead bricks. We also went past the user lab where the users can have a break while their experiment is running. So because the size of the x-rays is about the size of atoms and molecules we can tune the x-rays to tickle our samples in just the right sort of way that we want to. The really interesting thing with the synchrotron is that there are dozens of ways of using the x-ray beam lines to interact with the samples and I'm going to go into a couple of the simpler ones here. So this is your traditional broken bone x-ray using the absorption technique. You fire x-rays through the patient. The denser materials like your bones will stop the x-rays. The less dense materials like your flesh will stop the x-rays and you measure that differential and you can see the resulting image there. So that's a very standard x-ray procedure that you would use in medical circumstances. More complicated way of using the x-rays is to focus on their wave-like capabilities. So the once you find the x-rays at the sample they will go through the sample and the wavefront will get perturbed by that sample. Now because it's a wave and it's constantly moving if you give that wave that perturbed wave a little bit of time very small changes in that wave will get amplified further away from the sample. So you can see there on the right hand side there that the very small change that that sample has made in the wavefront has been amplified further on. So the trick there is to actually work out how far away your detector should be from the sample. And one of the interesting things is that x-rays get scattered in air. So often what the experiment experiment is trying to do is actually have their sample and their detector in a big vacuum box so that you don't get fuzziness from the x-rays being scattered by air particles. And this is an example of phase contrast being used on a very small bug and one of the things to note there is a very common feature that you get using phase contrast is small features. You'll often get a white fringe right next to a black fringe and that's a physical property of using phase contrast but it's also really interesting because it's one of the things that if you're using any sort of machine learning you can ask the machine learning to look for that change in colors and it will actually be able to work out a vector map of what you're looking at. This is creeped from the physics open lab which is a wonderful DIY physics website out there. So this is diffraction which is a very old technology but we're able to do it in really interesting ways at the synchrotron. Essentially what you're doing is firing x-rays at a crystal and the crystal in the case of the synchrotron is a microscopic crystal. It's not a big hug of diamonds as the image would show. Because of the regular structure of the crystal in the sample all of the electrons that hit it will all fire off at the same angles. So what you're hoping to do is work out what the really tiny crystal lattice layout is from the shadows that the x-rays produce. So if you have a look at that back panel there you can see some of the patterns that are occurring on this. So what you're really looking at here is all of the really small black dots that are forming the circles. So this is the shadows of the x-ray lattice that are being formed and if you rotate that sample enough times you will get enough information that you can work out what this 3D structure of the the crystal is at an atomic level. Another common medical use of the synchrotrons is computed tomography. This is where we use x-rays to take a very a lot of 2D slices of something and we stack them together using computers and we make a 3D model of what we're looking at. So that's lots of scans of the human brain and this is a completely separate image but one of the interesting things here is it actually shows the bed that the patient was lying on there also got imaged. So the the advantages of doing that at a synchrotron is that the x-rays that are produced at the synchrotron are generally much smaller than a medical beam line and they're much less scattered so you get much better pictures and much finer grained pictures. So it's there's this wonderful thing where the the chief tool that we're using are really tiny x-rays and the infrastructure around them tends to be rather large. I've mentioned that x-rays scatter in air so a lot of the experiments are actually done in big vacuum boxes so they're airtight boxes and they pump all of the air out and that means that inside that box the x-rays can't hit any air particles so they they don't scatter and you don't get any blurriness of the images that you normally do get on a medical x-ray. And those big boxes make everything a bit difficult because they take a long time to evacuate so when you're running an experiment in one of those beam lines you tend to not want to do one experiment you tend to want to do a hundred experiments because evacuating the box can take many hours. The accelerators to accelerate the electrons and to turn the electrons they use up a lot of power and they need a lot of cooling so a lot there's a lot of water cooling going on so there's a lot of pumping and there's a lot of plumbing going on and all of that tends to be on the large scale. In the filtering processes of each beam line we're generating a lot of x-rays at the start of the beam line and by the time that the beam is filtered and is hitting the samples there's generally a fairly small amount of it left. The rest of the x-ray beam line it gets shunted into blocks of material those blocks heat up and they have to be cooled and some of those water cooling is good enough but a lot of the other ones they actually have to be cooled with liquid nitrogen and such and there's a lot of work around that to make that happen effectively. The other the the other advantage of having a large storage ring is that we can run more beam lines off the circumference of it and we can run more experiments at the same time and one of the interesting things with the medical beam line at the synchrotron is the beam that is generated from the synchrotron is extremely parallel and if you wanted to put a patient in front of that you would have to move the patient left right and up and down a lot so the thing that they've done with the medical beam line is place the medical beam line outside the synchrotron building in a totally separate building so that the beam line has spread so that the patient only has to sit in one spot and the beam can go up and down and do a full scan and i've mentioned before that in order to protect from the x-rays a lot of the concrete and a lot of the lead has to be quite thick so like some of the concrete doors are like yay thick so you make anything that thick and it tends to be get big there's a lot of uses for the synchrotron medical microscopic samples looking at single cells and synchromolecules some cultural history stuff we can do interesting things at the samples as well so we can put the samples under strain we can pass radio waves microwaves through them we can heat them up we can do a lot of things to the samples that you can't do in other equipment there's a lot of work on battery technology at the moment and there's a lot of biology work so there's interesting cases where you can have a chemical reaction taking place and you can look at the chemical reaction as it's going so typically how it works is that months in advance researchers will submit their experimental proposal typically the synchrotron is way over subscribed so not everyone who has an idea for an experiment gets accepted if their idea has enough academic merit it gets accepted and then there's a long discussion with the beamline scientists the experiment wants to do one thing the beamline can't quite do that so they have to come to a middle ground and of course there's all of the communication issues where everyone thinks that they know exactly what's going on but nobody does we have on-site accommodation at the synchrotron because the synchrotron is a 24-hour facility when you arrive you'll get a lot of specific hours and it could be like three o'clock to six o'clock so you don't want to stay at a hotel off-site you want to be able to roll out a bed and show up to the synchrotron once that site depending on the type of samples they might do some chemical work or some physical work to prepare the samples for the synchrotron you run your experiment everything goes perfectly and the researchers get exactly the data that they were looking for the first time they don't have to rerun their experiments and then we as the scientific computing group we will keep all of that experimental data on-site as the canonical location of it for five ten years depending on which beamline and how much data you've got the samples can be pretty much anything we have specific holders for gases solids and liquids and as I previously mentioned we can do a lot of things to those samples so a lot of the experiments can be dynamic this is a co-flow holder the yellow is a glass capillary tube the blue is a buffer solution so just water and the the red pink is actually the liquids that is getting tested and we we are actually seeing the x-ray beamline go through that at the moment and the liquid is running through that capillary undergoing a chemical operation so we're actually able to image and scan a chemical reaction as it's happening and this is one of the physical holders that goes into one of the vacuum chambers so you don't want to run one experiment at a time you want to run a hundred experiments at a time if you can I was hired under the bright project the bright project essentially is looking to substantially increase the number of beams that the synchrotron can do at one point in time depends on how you count it but we're looking at eight new beam lines so very close to doubling the number of beam lines we're looking at standardising the hardware and software for the new beam lines the current beam lines they were all sort of produced by separate teams so all the hardware is kind of the same but quite different and all of the software is completely different for the new beam lines what we're trying to do is standardise on the hardware and the software so that if we improve the software on one beam line we can improve it for all of the beam lines and there's the bright plus project as well where what we're trying to do is once we make improvements to the software we can roll that back to all of the other beam lines as well and it's a really interesting software project because we've got those current beam lines there some of the beam lines are really happy to test out some of the new software that we're writing for them so that's really cool and when you think about the users of the software we're creating we have to be able to write software from first-time users who've never visited a synchrotron and they pretty much just want a big go button to chief scientists who not intimately know all of the details of their beam line and they want to tweak and test and try different things so there's a really wide range of disciplines that we've got to write software for. We have to be careful that we're writing for sleep-deprived people if you have a really bad experimental time slot of 3am you'll be sleepy so we have to write our software in such a way that it's safe and it can't break the equipment and there's different users as well a lot of the beam lines they have a very standard experimental setup where the sample is very standard and there's a lot of other beam lines where it is very specialised and the experimenter wants to do something very frunky and we have to be able to handle that. There's a lot of data that gets generated and for some of the current beam lines and some of the new beam lines if there's too much data for the researchers to take home so that data has to be stored on site which means that we have to provide the analysis tools on site and all of those tools must be made available for remote users as well. So the technology stack we're looking at we've got motors detectors and sample handlers as the hardware. We're using an open-source library called Epix and that's used in a lot of physical installations around the world so we definitely don't have to reinvent the wheel there. On top of that we're using PyEpix which exposes the C library to a Python world and then OFID which hides a lot of the details. Blue Sky we're using to orchestrate our experiments and then we'll be using creating web frontends to drive all this and that means that they're available remotely as well. So I'll skip over Epix, it's a low-level networking tool and I'll skip over PyEpix because it's a very good wrapper around Epix but it doesn't provide anything in particular. OFID is the part where we really come in. It abstracts away all of the funky things that each of the hardware items does so if there are any quirks in the hardware we have to encapsulate it in the OFID layer. Blue Sky is our orchestration tool that we'll be using and it basically you come up with a recipe for how to run an experiments and it handles all of the problem cases that you get where you've halfway through an experiment something's gone wrong you can pause the experiment fix it up and continue on and it's in use at other synchrotrons around the world so we can build on a lot of that. It is an asynchronous main loop based infrastructure and it handles a lot of the issues that we would see. A basic experimental run like this is where so this is a no-op experimental run. The chief context that gets passed around is a document and all of the data and metadata that all of the samples and all of the detectors collect go into the document. So a simple experimental run here is where we move one motor to maybe spin the sample in the BAM line and then we have a detector off the side detecting the bounced x-rays. And as we go along we accumulate all of that data and metadata and it goes into the final result. But the really nice thing with Blue Sky is it lets us do multiple things at once. So we might have the same experiment as the last time but we also might want to monitor the BAM line. So some of the experiments they might want to change the strength of the BAM line that's being fired at the sample and we can do that. And at the same time they might want to be heating the sample and they might want to be remembering exactly how hot the sample was. And because Blue Sky is asynchronous we can kick off all of those things at once and everything will orchestrate perfectly in the end. And all of that metadata ends up in the final experiment documents. I'm not much of a gooey person but I think we're in a really interesting time at the moment where you can target the web at HTML and JavaScript or WebAssembly if you need that. And things like PyOd died and things like the QT WebAssembly target. I think that's really interesting. And if you're not targeting one of those two gooey back-ends I really hope your hardware is very funky. So that is the main content of the talk. So thank you very much for that. Does anyone have any questions? If you have questions please use the microphones. If you don't have any questions I'll go on to some of the after-credits scenes and please come up and interrupt me. Friends around the world, where does the Australian one fit into the big picture? Like is it powerful? Is it a smaller one? Is it? It's relatively small but we do have some interesting stuff going on in particular places. So we only have I think, don't quote me, I think we can have 36 beam lines if we use up all of our space. The bright project I think in the end will probably have 18 or 19 experiments that we can run at the same time. A big thing with having a synchrotron in a place like Australia is that it means you don't have to get in a plane for 24 hours and go to the States or Singapore or something like that. There's a lot of interesting science being done there but I think it's fair to say that we run on a shoestring budget. Hi, one more question. You mentioned that there is another synchrotrons that run the same software. Can you estimate the amount of them? I cannot. The one that I know runs it is NLS2, which is the national light source in the US and they are a big synchrotron. I can walk around our synchrotron, you need to get into an electric golf cart to go around their synchrotron. I think there's maybe two or three others using the blue sky suite of tools. Okay, thanks. So this is one of the interesting use cases of microbeam radiation therapy. So you've got cancer and you do chemo and you can do radiation therapy. Normally, when you're getting radiation therapy, the bit of radiation is like a square of radiation. That's maybe 0.5 centimetre squared and it's targeted at your tumour and you get rotated through that because they don't want to radiate all of your healthy cells in front of that tumour. Or they have to, but when they're rotating it, they minimise that. This is something that you can do with synchrotron radiation where instead of having a square block of radiation, you have spikes of radiation. So you actually get hit with a bunch of lines of radiation and if you can see on the right, you can see the lines going through the tumour and it turns out for reasons that we don't know yet that by hitting tumours with that style of radiation, we can amp up the radiation to really high levels. So instead of going back and having four or five radiation treatments, you can have one radiation treatment and it knocks out the tumour in one. And for some reason, doing it in those lines means that it doesn't hurt the healthy cells that it goes through. And there's ongoing research at the synchrotron and at the Japanese synchrotron as well to see if we can turn this into a viable cancer treatment. And I believe they're going to start working on animals at the Australian Synchrotron in a year or two as test subjects. So animals who've got cancer that there's no other treatment for. And no one really understands how or why it works yet. But being able to go in and have a single dose of radiation on cancer that will knock out the tumour and not affect your cells, essentially amazing. So this is another medical use of synchrotron radiation where one of the problems with radiology at the moment for mammograms, if the patient has high dense memory glands, the standard radiology cannot differentiate between cancer and dense flesh. It's a known problem around the world, but there are no good solutions for it. But this is a potential solution for that using synchrotron radiation where the radiation can be tuned so that it can pass through the denser flesh and spot cancer problems. This is... So the previous example you just gave, where the synchrotron produces higher quality radiation or more configurable radiation that can be used to detect things that a regular standard machine can't. What is the end game that you would bring people to a synchrotron for the therapy or you would modify existing machines to be more? Yeah, so we have all of the equipment to be able to have patients at the synchrotron. At the moment, it's very much in a... So the interesting thing with the Medical B mine is that it is outside the synchrotron building. So it's already in its own separate building. So it's feasible that we could have people come into the separate building and treat them as patients and they wouldn't need to have someone like me explain what a synchrotron is at the front. And we're a while off being able to do some of these medical things but that is definitely the end goal. So everyone, you can leave for lunch now. I've got like two other pretty pictures that you're welcome to stay but if you're very hungry, that's totally fine as well. So this is one of the examples of diffraction that was done at synchrotron and what they were actually able to do for the first time was to show the physical process of the malaria parasite getting into the red blood cell. Obviously what they're looking at here is incredibly tiny and you really need very finely tuned very small X-rays that at the moment, the synchrotron is the only way of doing that. And the goal is if we can work out exactly how the malaria parasites are getting into the red blood cells, then we might be able to counter it. So the first thing that you've got to do is know your enemy. This on the left is a painting that was purchased by the National Gallery of Victoria. Lots of art critics like myself sort of looked at it and said it was a little bit ugly. You can see on the right half of her face, it's rather black and there was lots of guessing about what was going on and using the synchrotron, I'll just try and find my mouse, what they discovered and you can see there's an old painting and you can see the face in that. And what the artist has done is he was working on that first painting. At some point in time decided to give it up. Instead of painting white over the top, he's painted black over the top, turned the painting around and stuck a new painting on it. So there's a lot of uses for the synchrotron for cultural heritage stuff. So it's really common now when you're dealing with mummies, for example, that you're not actually opening up the mummies. All you're doing is really high quality CT scans and looking at things and all of this is non-destructive. This was another case, it's a medical case but it's for animals. They were able to have a look at the physical shape of the PCD virus, which is beacon feathers for a endangered Australian bird. And the hope is by being able to physically look at the structure of the virus, they'd be able to create a vaccine for it. And I think this is the last one. So what they're trying to do here is breed grains with higher levels of micronutrients and in this case iron, copper, zinc, potassium, calcium and manganese. And the idea here is that they wanted a quick way of being able to, if they wanted to find out what breeds currently had good qualities of those micronutrients, they wanted to crossbreed them and then they wanted to be able to see if those crossbreeds had traits of the parents. So there's really, so the examples that I've chosen there, I had to choose the ones that had really good pictures. There's a lot of other examples of synchrotron use that don't have good pictures, so they don't work well in this sort of slide. But I really just wanted to get across that synchrotron radiation can be used for a really wide variety of research fields. And that is the end of my presentation.