Chris Yip 0:01 Welcome to Tell Me More: Coffee with Chris Yip, the official podcast of the Faculty of Applied Science and Engineering at the University of Toronto. In our second season, I want to focus on the journey, how people got to school, what they did during their time here, and where they've ended up after graduation. You will meet students, professors and alumni, and learn what places them at the heart of designing bold solutions for a better world. My guest today is Professor Dionne Aleman from our Department of Mechanical and Industrial Engineering. Dionne and her team apply tools from the worlds of operations research and machine learning to improve medical and healthcare decision making. Their work has included some pandemic modeling. She is also deeply invested in improving our student experience serving as Associate Dean of cross disciplinary programs, as well as a co-chair of our Engineering Positive Space Committee. Professor Aleman, welcome to the podcast. Dionne Aleman 0:58 Thanks, Chris. It's great to be here. Chris Yip 1:00 I'm looking forward to our conversation. It's going to be wide-ranging, right? Got to talk a little bit of research, got to talk about some of the stuff you're doing in the cross-disciplinary space, some of the stuff you're doing in the outreach space, and I think we'll probably get a little diverted along the way so be prepared. We always start this with kind of the standard question. Tell us about how you grew up and when did you realize that engineering was for you? Dionne Aleman 1:25 I actually didn't know I wanted to be an engineer for really quite a long time, all the way up until the moment that I was applying for universities. I always liked you know, math and science, but I also really like English and journalism. And for a while, actually, I thought I wanted to be a high school English teacher. At my high school, there was this process wherein every junior, so 11th grade, had to meet with the school guidance counselor to talk about, you know, career plans, university plans, and she was like, well, what do you like to do? And I was like, well, I like math and I like science but I also like, reading and writing and...and she's like, I just learned about this type of engineering that it sounds might be really interesting to you. It's called industrial engineering and it's all about using math to do things more efficiently and it involves a lot of communication and working with other people and a lot of teamwork and you've been heavily involved in sports. I was two-letterman - two-letterperson varsity athlete, and I was like, yeah, you know, actually, that actually does sound really great. I really do like being efficient and finding the fastest way to do things and with the least amount of effort, because Oh, my God, my whole family has always been like, do things faster. Why are you spending so much time doing that? If you did this way, you'd be done already and I was like, Wow, that just really speaks to me. And I went home and I was telling my parents over dinner, oh, Mom, Dad, I think I found the exact right career for me. It's this thing called industrial engineering and I started telling them about it thinking they've never heard of it. And my parents both looked at me, like I was insane. And they're like, are you crazy? We're both industrial engineers. How did you not know? Like your uncle has a master's in operations research. Like half my family's industrial engineers, but I'd never heard this term before because nobody has a job title, quote, industrial engineer, unquote. Like my uncle with the master's, he's a hospital CEO, my dad was an executive VP in the county transit system, my mom ran our family business, nobody has the word industrial engineer as their job title so you just don't hear it, right? So I almost changed my major to English right then and there. I'm like, no, I refuse to be the same as everyone else in my family. But well, I stayed in industrial engineering. Chris Yip 3:39 You're right. Industrial engineering is one of these fields where people - it's everywhere and yet people don't see it as much. It's a bit of a mysterious field but as you pointed out, it's one of these fields, which is so important. When I heard the term operations research, I'm thinking hospital operation (laughing) but I don't think that's totally what it is, right? So for our listeners, could you explain what operations research really is? And really, where is it typically used? Dionne Aleman 4:06 Yeah, operations research is really just a terrible, vague term. It's sort of been taken over in the popular lexicon via the word analytics, which you know, has a very trendy, popular feel to it and people sort of feel like they kind of intuitively understand what is meant by analytics, although operations research, I would say is quite a bit broader than what you might think of when you think of analytics but OR gets its name from military operations. So operations research as a field basically began during World War Two. So it's as far as engineering disciplines or really fields of study, period, it's really quite young. If you've ever read anything about military movements, military operations, there's a lot of logistics involved. Lots of people lots of things have to be moved around have to be coordinated getting things to remote locations and getting them back out again, and how do you do that? And operations research was born out of the need to be very efficient and economical with moving both troops and supplies, like supply chain management, during World War Two, and so military operations eventually, and really quite quickly came to realize, wait, this is not just good for military, it's also good for just general businesses and products and moving things around and so the word military just kind of got dropped off and then we were left with just operations. Chris Yip 5:23 So it really is critical to all aspects of life right now. I mean, supply chain, as you talked about, global supply chains, as we've been dealing with, basically getting goods, products, production, processes, people everything movement, and identifying where things get stalled, I guess, to borrow a term of the...(Dionne: Bottleneck) a lot of bottlenecks...(Dionne: It is what we call them) and how you de-bottleneck the process. I use the analogy of operations thinking of surgical operations, but your research is really focused, kind of in the healthcare space. Give me some context, what area are you focused in? Dionne Aleman 5:59 So at the moment, my primary areas of research are in pandemic modeling and planning, really looking at a lot of different facets from simulation models to predict the trajectory of COVID to helping predict the severity of any one person's COVID experience at the time of PCR testing to vaccine prioritization. I also do a lot of work in organ transplants, trying to be more efficient and provide more accurate predictions on success rates. So looking at bone marrow transplants, I'm looking at kidney transplants and creating large chains and cycles of one person donates to another who has someone who donates to another who has someone to donate to another, to try to very efficiently get lots of people the kidneys that they need. And we're just right now branching into liver transplantation as well, looking to see, can we predict how long a person might survive with a particular transplanted liver before needing another transplant. And then I also do a lot of work in designing various types of radiation therapy treatment plans for cancer treatment, which is work that I started during my PhD. It was actually something that really attracted me to stay for my PhD learning that I can use my mathematical skills to actually help people as opposed to just you know, help businesses make more money, which is kind of where I always thought I was going to end up. I thought, wow, I could use my knowledge to help treat cancer, to help improve the quality of people's lives with respect to their health, and that really spoke to me and so I did that for my PhD and, well, I'm still doing it to this very day. Chris Yip 7:29 So in the context of the transplants, is the focus of your research, really around the lifespan of the transplant, but also the kind of the biology of it and sort of the supply - sounds a bit weird - the supply of things like liver lobes and whether someone's living donor of a liver or whether you're getting it from a death? For instance, when you're transferring that kind of thing, how do you model all of those different parameters? Because it seems that would be incredibly complex. Dionne Aleman 7:57 So it depends on the particular problem we're looking at. And so like with kidney transplantation, a lot of the organs that are received are from living donors, right? Somebody donates a kidney. And you know, can we spark a big you know, cycle of donations with with one person's donation is essentially what what we're looking at. But interestingly, a lot of times, somebody will just choose to donate a kidney not because they have, you know, a loved one that needs a kidney, they just decide, I'm not doing anything with this kidney, somebody else could make a lot better use of it. And so, which I think is just a truly inspirational thing to do. And then wow, so now we basically just have like a free kidney, like, what can we do with this? How can we really maximize the value of this one kidney? If we give this one kidney to this person, we can give their donor's kidney to this person, their donor's kidney to this person and I really create a lot of opportunities. And so while most of our work is focused on the donated organ just appears, okay, now, what do we do? We are starting to look into analyzing how often do these sorts of altruistic donations appear and if our models are used, like on a consistent basis over the course of a year or two years, how many more kidneys might get transplanted compared to the status quo. For liver transplants, so you mentioned living donations for livers so that is certainly possible. Apparently, it's actually quite rare at least here in Canada. I was just learning from our collaborators just last week, as this is a new project, we're still sort of gathering the foundational background material, most liver donations are actually deceased donations. So basically, you get the liver that you get, and you've got to see who on the waitlist is going to be a good fit for that liver. And that's actually really a very interesting situation that plays a lot into equity and fairness issues because lots of people on the liver transplant are sort of fundamentally structurally disadvantaged, specifically women, because one of the big factors in liver donation is literally just the physical size of the organ. You know, how's it going to fit into the patient and as it turns out, most The origin to end up being donated are male, and most male livers are too big for for female patients. And so that, you know, that ends up with this big problem of women end up being on this waitlist for a lot longer than than men, right? And is it because every single liver that comes along is not appropriate for any of the women on the list? Or maybe you know, like a small enough liver comes along, but it's a better match for for a male candidate and so it gets fit there but then that does that create these structural inequities. So that's something that I'm interested in examining and that problem as we also look to see, can we predict the lifespan of that donated organ before before the patient might be in another organ? And you asked about biology. I will say, I don't know a ton about biology. I learned basically what's needed for each particular problem and we always work very closely with clinical collaborators like we have to have actual transplant specialists who work with us who tell us what's important, how do things work, what do they care about, and to be able to validate our approaches and make sure that we're not doing anything that's just totally off the wall, and somewhere to look at the numbers has taken us into bizarre direction. Chris Yip 11:05 Let me divert a little bit back to the pandemic models and ask a quick question. What do you think as someone who models pandemic behavior, and just the whole thing that's been going on now for the last two, just over two years, what should the public know and understand about models in general, or pandemic models in particular? Dionne Aleman 11:23 There's a famous quote by George Box, a famous statistician, that says, "All models are wrong, some models are useful." And this is the kind of mantra that that we live by when the rubber hits the road in these operations research industrial engineering modeling applications. A model, you know, it's only a model, right? You try to take into account all the big ticket items, everything that seems impactful everything that is, let's say, within the control of the decision makers, in this case, your public health officials, politicians, the levers that they can pull, and you're based on what you've seen historically, various things going on around the world, you try to make an extrapolation. And how good are these models? Well, when it comes to pandemics, it is literally 100% impossible to ever know, because most models are modeling what's happening right now, right? And if we do nothing, oh, look at this giant exponential spike up, it's going to be bad. Public officials do something, they implement some sort of mitigation measures, not necessarily full lockdown, but just continuing to tell people to wear masks, capacity restrictions or something like that, and then, we don't see that big exponential spike, maybe it's smaller, or maybe it's just a little hump and well, does that mean the model was wrong? How can we know? Right? Because the situation changed. The types of models that I developed, they're called agent based simulation models, like I do try to, you know, model an entire population, every person individually and then we look at, okay, what happens if we shut down like this sector or that sector? What happens if this many people get vaccinated in these age groups of vaccines of this efficacy? And this and that, or whatever, and try to see what happens, even going so far as to include individuals comorbidities based on the regional prevalence and what that might mean for hospitalization rates. And so trying to get very, very accurate, but even still, even if the public health officials that I'm working with implement the policies that I say the people don't always adhere to the policies exactly, as they've been prescribed, and a lot of times we will look at it adhere to what if 50% of people do what they're asked or 75% or 90%, and try to make guesses. But you know, at the end of the day, like, even if the model is perfect, which it never ever is, right? All right it's good, right? Models, good models are good enough to make decisions off of but they're not crystal ball clerk clairvoyant, you know, soothsayers with winterlike eyes into the future to see what's going to happen. Even if the model is perfect, like I said, even though we know it's not, people aren't going to behave exactly as expected, like one person might go out to dinner and that might spark a huge outbreak, just a totally unfortunate random happening of chance and you can't predict, you know, all of these possible outcomes. And even in my models, where, you know, I'll run like 500, 1000 simulations to really see what's happening at the tails of the distribution, like the really unlikely events like those like nuclear super spreader events, oh, that's only happening like 2% of the time, but this policy versus 20% of the time that that policy, even if you go with the with the safer, more conservative policy, you still might end up with just essentially really bad flips of the coin that end up with something bad happening. And it doesn't mean that the models are bad or wrong, it just means that we can't predict the future, we can just say, well, looking at what's likely to happen, here's what we should probably do and how we should probably try to respond. Chris Yip 14:49 Fundamentally, that's a really important point, right? Is that models at some level, they're as good as the data you put into them. They're as good as the sort of parameters you adjust in tune by it and I'm not sure even in your models now, I mean, are you able to account for what we've seen with, you know, Delta and Omicron, and all these different variants? How does that affect your models? Dionne Aleman 15:13 So well, my models do, but your one limitation is, at least at the moment, is that they only consider one strain circulating at a time and that's something that I was meant to expand out, but ultimately, Omicron came, washed over everyone like so quickly, it sort of became a bit irrelevant. But yeah, my model is in effect, really, most disease [unintelligible], you can't tune the virulence, how contagious something is, how long are people contagious when they're contagious, how many people are asymptomatic because if you don't even know you're sick, you don't know that you should stay home and so you just go out and just spread your germs everywhere. The bigger challenge is not so much the different strains, but knowing what's the ground truth, like what's actually happening right now. How many people have actually been infected with Omicron, let's say in the past, like, four months, and therefore are ostensibly immune from catching it again, in the next couple of months. Reinfections are real, they happen, right? But there is, you know, a window of a grace period where it's very unlikely, not impossible, just unlikely. But we don't even know, right? We don't know how many people are actually infected, because testing has been so dramatically reduced, not just here in Ontario, but across the board. Of course, here in Ontario, there are a lot of like, difficult hoops and rules that have to be followed to get tested in the first place, which you know, puts people off getting tested. And thankfully, now we have wastewater data that we can look at to try to get a handle of what's really going on. But it is difficult to match wastewater data to an actual number in the population, like how many people is that? Like 100,000? Like, I mean, you've seen the ranges from the Ontario Science table, like at the peaks of the wastewater, they were thinking like 100,0000-130,000 a day becoming infected, that's actually like a really big spread. You know, if you think over the course of just three or four days, that's an extra 100,000 people just in the in the error range there. Information like that can really change what a model predicts because if 80 or 90% of the population has been recently infected and you know, some other amounts may be not affected, but did get immunity from the vaccines even if COVID continues to spread, it'll fizzle out on its own very quickly. But there is like an inflection point, where if not enough, people have had it, there's just not enough prevailing immunity, right? And I'm not going so far as to say herd immunity, but just immunity that things will just sort of die off more or less on their own. We don't know where that is, right? Because we don't know how many people have been affected. So the less information you have about your ground truth, the less accurate you can be with your models. Chris Yip 17:36 This whole idea of operations in the healthcare space is a great example of something that you're really tightly interwoven with now, which is your role as associate dean for cross-disciplinary programs, could you give our listeners a bit of a sense of what this is, what is our cross disciplinary programs office and your role in it? Dionne Aleman 17:58 The cross-disciplinary programs office is all about basically sticking engineering where people might not necessarily realize it belongs. It sort of encapsulates my entire research. So you know, when we say cross disciplinary, we're basically just talking about reaching across disciplines. So in the CDPO, we're responsible for running and creating all of the minors and certificates that we offer to our engineering students. So things like minors and certificates in AI, in business, in bioengineering, in music, we have just such an incredible spread and variety of topics and you can choose to concentrate a little bit in a topic and get a certificate or a lot and get a whole minor and it's all about just - to borrow a political phrase - for reaching across the aisle, so reaching across doors to our colleagues, not just across the University but even within engineering from one department to another, to say what we think our students in this program would really be interested in having an opportunity to get a minor in, like, let's say robotics, right? That could draw students from mech and from ECE, you know, what can we do to create a unique learning opportunity for our students so that they can really put themselves in a really interesting, cutting edge career field right off the bat as they graduate? And we do the same in the Faculty of Arts and Science. But just recognizing that engineering has a lot to offer, a lot of different disciplines and that our students have a lot to offer just beyond formulas and equations and chemicals and everything else that they learn about in engineering, they have broader interests. We train our engineers to think in in a systematic way, to think very broadly about the way the world is interconnected, of course, always through the lens of their own particular discipline but how can we expand that out? One thing that we have launching this next academic school year, so starting in September is a certificate in public health. Chris Yip 19:57 Oh! Timely (laughing) Dionne Aleman 19:57 Very timely (laughing) but of course public health is not just about pandemics, right? There's all sorts of things like, you know, environmental health and occupational health and safety. But this is something that's...the interest is born out of current events, right? So I think one of the key opportunities in our office is looking like what's going on in the world? Is there an opportunity or an interest in our students to contribute in that area? And how can we build educational opportunities for them to really just hit the ground running in your very niche intersecting areas? Chris Yip 20:32 Yeah, we're really excited I think about this public health one. I was reminded when I think when we approached the dean there that he reminded us that sanitation engineering was for the Foundations of Public Health. Back in the day it was the design of sanitary sanitation systems, which was largely bringing in civil engineering, and works both ways, right? Because I think we're also supporting public health students in public health, partnering and working with our design teams as well so I think it's going to be a great partnership going forward. So a couple other things I want us to avert away from our talk about operations research, and you're an active member and co-chair of our Engineering Positive Space Committee, give us a little bit of a sense of what that committee is working on and its mandate. Dionne Aleman 21:13 Yeah, so Engineering Positive Space is a LGBTQ+ advocacy group within engineering, and it's all kind of just a bit ad hoc and informal. It actually hasn't been around all that long, and maybe just 10 or 12 years, something like that. It's basically a place where members of the LGBTQ community in engineering, and well, really anyone who wants to join us, we're not picky, can just come in and just talk about issues that they see around campus. Is there anything that the committee can do? Or have there been incidents of concern? And how can the committee help respond to those incidents, and try to create a more welcoming and inviting atmosphere for our LGBTQ population? And that's not just students, but also faculty and staff, Chris Yip 22:01 In engineering we've been working hard to make sure that we're an inclusive and welcoming community. Of course, there's always lots of work to continue to do and June is pride month so what do you want our listeners to know about sort of positive space that relates to engineering or the Faculty? Dionne Aleman 22:17 Well, we have a student group - QueerSphere - that is basically the focal point for actual activities. So you can check out their website on Skule, sign up for their listserv. We also have at the university level, positive space, and also the sexual gender diversity office, both of which arrange events or at least distribute a lot of information about events going on related to pride and in Engineering Positive Space, we often redistribute those events. Blue and Gold, the student group they often do a big float for the Pride Parade. So obviously, that's been on a little bit of a pause during these COVID times. I'm hoping that I think it is back also. So I'm really excited for that because it's always it's fun to march in the parade, it's fun to see the floats and all the students, it's a it's a good old time. Chris Yip 23:08 So there's always these sort of off the wall questions that I like to ask. I forget what started this so we started talking about cars for some reason, right? Dionne Aleman 23:16 I think it was because someone had tried to steal my car or my wife's car (Chris: Oh, that's right) in the middle of the night (laughing). Chris Yip 23:23 Yeah, we started talking about that. (Dionne: And I mentioned it to you) Yeah, we started talking about whose cars were easier to steal and the types of cars you would steal, and off we went on a tangent. So tell our listeners, because this part of it knowing our faculty and the members of our community but you are a bit of a car fiend, I guess. Dionne Aleman 23:43 Yeah, probably unfortunately accurate. Chris Yip 23:47 Can you give our listeners a little bit of a sense of what have you to your vehicles as it were? Dionne Aleman 23:52 I have a hot rod and I won't say exactly what because should I ever manage to find an insurance company that will insure her I don't want to be recognized driving around town because I've never seen another like her in Toronto. But it's an American muscle car. And wow, there was a time in my life where every dollar that I earned went right into my car. And I love her more than anything. I'm pretty sure I love my family more now. My car was like the number one reason for my existence. I've got a tool chest that's taller than I am. I've done pretty much all the mods that you can do and I was actually right about to do stage two heads and cams, that's what really gives your car that kind of [Dionne makes car revving noise] sound and just take off the line - the track. But then it turned out I was graduating with my PhD and getting a J O B. I was like okay, maybe I better put this on hold until things are a little bit more stable, tenure and all that stuff because once I start doing this, what if it takes me a long time especially because I don't have like a good mechanic that I know here in Toronto in case things go south. And I need to bring in some, some bigger brain power or bigger, bigger tools to help me out. I was like, let me just put everything on pause, timeout. But then it turned out after we moved here, I couldn't find an insurance company that would like touch my car with a 10 foot pole. So, so she just sits in a garage being a garage queen. Every now and then I go on, I'd look at her, sit inside, maybe start her up (laughing). Chris Yip 25:27 I was going to say maybe turn it over just to make sure it still runs. Dionne Aleman 25:30 It's like Ferris Bueller's Day Off. Like Cameron and his dad's Ferrari, just go and just look at it. Chris Yip 25:36 So here's a question for you though. I mean, you've got obviously an internal combustion car and then you've got EVs coming out, electrical engineers are in partnership now with Porsche around EV technology and training their staff on the new technologies coming down for electric vehicles. But they're also making a move, right? So their only gas powered vehicle will be I think, a 911. All the rest of their fleet is moving towards EVs. So what are your thoughts on that regard? Dionne Aleman 26:05 I think it's amazing. Oh, I think it's amazing. The power that can be had with electric vehicles like it's awesome. Like, don't get me wrong, there's nothing like, like, just the feel of like a cam rolling around the engine, like so hard that it just like shakes your car a little bit, like you're idling, and it's just like [Dionne makes car revving noise] like you just feel it like, I love that. (Chris: But...) I don't think EVs are ever going to replace that but at the same time, I mean, look at the power output of the EVs like you can't ignore the numbers that are coming out. Chris Yip 26:36 Dionne, I want to thank you. This has been an amazing journey. Thank you so much for giving us an insight and some of your time today to talk about this. Dionne Aleman 26:45 Chris, it's been great. A lot of fun. Chris Yip 26:51 Thanks again for listening to Coffee with Chris Yip. If you want to catch up on past episodes, or make sure that you don't miss the next one, please subscribe. We're on Apple Podcasts, Spotify, and more. Just look for Coffee with Chris Yip. You can also check out @UofTengineering on Twitter, Facebook, Instagram and LinkedIn for more stories about how our community is building a better world. And finally, if you'd be inspired to join us, we'd love to welcome you. Whether you're thinking of taking a degree or working with us on our research project, you can find us online at engineering.utoronto.ca. Or you can visit our beautiful campus in Toronto, Ontario, Canada. I hope I can join you for coffee soon.