MIT Climate Engagement Forum: Research Collaborations to Decarbonize the Energy System
[Bob Armstrong] Good morning. I'm Bob Armstrong, Director of the MIT Energy Initiative. And I'd like to welcome you to this first of two forums on engagement.
Both this symposium and the next one, the next one hosted by Professor John Fernandez of the Environmental Solutions Initiative at MIT, are being held under the auspices of the Office of the Vice President of Research at MIT, Maria Zuber. With these events, Maria Zuber is continuing MIT's initiatives, such as MIT's six climate action symposia to engage community members around urgent climate and energy issues. You'll hear closing remarks today from Vice President Zuber. Today's symposium looks at engagement with outside organizations, industry in particular, on climate and energy research. With vignettes of current and recent engagement activities, we're looking to share a small handful of examples of how working with industry has catalyzed progress in the electric power sector, lifecycle analysis to inform decarbonization efforts, and fusion energy, to name a few. You'll hear from industry representatives, investors, faculty members, and alumni who've been involved with these projects from different angles.
Something you won't hear as much about today but another kind of engagement that's very important to us is informing public policy through our multidisciplinary, multiyear studies, involving teams of faculty, researchers, and students from across the Institute. The most recent of these, the future of energy storage, is currently underway. And when it's completed, our researchers will present findings to policymakers in DC and across the country, as we've done in past studies, including those on personal mobility, nuclear, solar energy, and the electric power systems. While we're trying to pack a lot of information into a pretty tight schedule this morning, we'll have some time for audience questions and answers. So I ask you to scroll down on your screen and enter your questions into the Slido field under the livestream video.
Thank you very much for being here. It's wonderful to have so many community members who care deeply about MIT's climate and energy commitments. We're grateful, certainly, to gather virtually at this time. And we look forward to the time, hopefully not so far in the distant future, when we can make these gatherings in person.
Now, I'd like to introduce Professor Angela Belcher, who will moderate the first panel, the first session. She is the James Mason Crafts Professor of Biological Engineering and Material Science and Engineering. She's also head of the Department of Biological Engineering, does a broad range of research in biomaterials, organic interface engineering, synthetic biology.
And she's, just generally, a wonderful scholar, educator, and innovator, entrepreneur. Angie, I'll turn it over to you. [Angela Belcher] Thank you, Bob, for that very kind introduction. I'm very excited to welcome everyone today for these panels. I think it's going to be very exciting.
We're going to have three different topics. And the part that I'm excited about is it's bringing together industrial collaborators with academic collaborators at MIT. And having done this myself for over 15 years at MIT, I really understand the importance of the close intersection and the close collaboration with our researchers and academics at MIT with industrial collaborators with the idea of really making an impact, really moving the needle in these particular areas.
So I'd like to introduce the panelists in our first topic, which is renewable energy, machine learning for predicting the longevity of offshore wind turbines. We're going to start by having about a 10 minute discussion. And we'll have questions at the end of each of the three panels. And so our first is Sofia Koukoura who's a senior performance analyst engineer at Scottish Power and Kalyan, I'm sorry Vermesh-- I'm sorry, Kalyan, who's the principal research scientist at MIT lab for information and decision systems. So I'd like to start by asking Kalyan a question on could you tell us a little bit more about machine learning for this project? [Kalyan Veeramachaneni] Thank you, Angela.
So in this project, we're using machine learning to predict failures for wind turbines, failures of components, like drive train, or controller, or converter, and items like that. And what that allows for Iberdrola is as we predict these failures far ahead of time, they can schedule repairs. They can pre-replace the components that are likely to fail. And as a result, increase the uptime of the turbines.
I think, increase the availability of renewable energy, especially from wind turbines. What was really surprising to me, when I started this research several years ago, actually, a little bit into the data-driven technology, that turbines do record a lot of data. They record data at a millisecond level.
And they record data about vibrations, about the power they're generating, about the heat, temperature, and pressure, and several other readings. And these add up to terabytes of data that Iberdrola is recording and monitoring through our wind farm. And then there are several wind farms. So one of the biggest challenges that we are working right now is not only to predict the failure for one turbine, but also how to scale up this technology to use across farms in Iberdrola's portfolio.
So that's the real exciting part about this project. [Angela Belcher] Well, that does sound really exciting, in terms of the scaling. Sofia, could you talk a little bit about how this type of work is valuable to you, and how working with MIT has been, and what the collaboration is like, and what this approach and technology brings to the table? [Sofia Koukoura] Yes, absolutely. Thank you, Angela. So this project basically enable us to have a machine learning solutions for our predictive maintenance for all our renewable energy assets. So one thing that is really good about this project is that it enables us to do this algorithm development dynamically.
So we don't get a black box solution by the end of the project, but we are actually building this with the researchers, which is really, really good for us. So we can tailor made to our needs. And we are actually seeing the code. It's fully transparent. And we're being able to contribute to it. So that's a really, really useful exercise for us.
And we're learning a lot with the researchers. The other really, really important thing is that, as Kalyan mentioned, the solutions being developed are very flexible, and reproducible, and scalable. So in a worldwide context, we know that we are in a climate emergency. So we need to decarbonize the energy sector. And that means more and more renewable installed capacity.
And in terms of Iberdrola's and Scottish Power, so Iberdrola is a parent company. Scottish Power is part of it. So it keeps on growing and getting more and more assets, that is wind turbine assets, solar assets, batteries. So that means more and more data from different types of machines, different types of technologies. Even when we talk about just wind turbines, there are so many different model data types. So the solutions that we are getting from this MIT project are really flexible and can be incorporated in different types of assets.
And that's really, really useful for us. And as a person who has worked in academia for many years before, in the field of machine learning actually, I can see that bridging the gap between development and deployment of a product is a very big leap. And the team at MIT are really helping us to actually bridging this gap. Because when I started doing my PhD, I trained one machine learning model. And I got a 99% accuracy.
And I thought, OK, I solved the problems of wind turbines. I solved climate change, that's perfect. But obviously, as you get into logistics of data science, you understand that there are much more challenges in actually deploying those models.
So I'm really happy to be working with the guys. [Angela Belcher] That's great. Thank you for sharing that. So how is machine learning, Kalyan, how is machine learning, that you've been working on this project, how does it relate back to the other research that you're doing at MIT? So has this connection expanded your research in other areas? [Kalyan Veeramachaneni] Very significantly.
I think, as Sofia pointed out, I've been doing machine learning for almost 12 years now, applied machine learning, not just theoretical machine learning. And what has changed is, historically, we used to get, oh, here's the data, and we'll go and do something with the data, we'll write a script. And we'll try to make a presentation of a report.
We can predict this. Or we could do this and things like that. And that's where it used to stop.
And right around 2015 or '16, things start to switch around. And we start to switch as well. So as we start to make much more open source software. We figured out how-- if you want to get buy in and machine learning be used, ultimately, to predict failures or to have an actionable outcomes, we have to have buy-in from the domain experts early on in the process.
So now everything has flipped, which is we now build open source software systems. And we figured out how, at the right places, you would have to get input from the domain experts, people who are on the ground, like Sofia, who know the domain much better than I know. And how do we get them? So we don't want to overwhelm them with software, engineering, and platforms, and code. But we want to figure out how do we give them right interfaces so they can input their knowledge that actually enables the machine learning to do better. And then as a result, we also get the buy-in from them, because, ultimately, they have produced that result. They are more likely to use it to take actions.
So that has informed us throughout our projects. So almost all machine learning, applied machine learning projects, right now we do, we actually have that sort of interaction with the domain experts. And we've also started to find that it's the domain experts, ultimately, people who are on the ground and are responsible for taking actions, those are the right people to interact with if you wanted to take machine learning to sort of deliver value to the society. [Angela Belcher] So that that's actually kind of a perfect lead in to my next question, which is to both of you. Are we closer to the promises of machine learning? Are we seeing real value that people have been hoping for? [Kalyan Veeramachaneni] And I'll answer that. And then I'll let Sofia answer.
I think, broadly, yes, that has been a challenge for machine learning community, to deliver value, ultimately. And values should be measured by actions taken in the field and outcomes delivered. So in this case, it would be how much availability do we have as a result of this project, availability of renewable energy power on a daily basis? That has been a significant challenge for almost all machine learning community, to deliver the promise that we have come to tell everyone. So projects like these have driven it closer. So we are almost there. We have been working for a year, year and a half, on this project, and I'm really excited to showcase that how much uptime do we have because of this project.
So I'll let Sofia speak for that, to her side as well. [Sofia Koukoura] Thank you, Kalyan. Well, I think the project is, now, just the start. And it's up to us to make the best out of these models. And, basically, how to use the outputs of machine learning to make decisions, that's now the most important thing.
So we're looking forward to actually test it, and see how much is there the cost reduction and the availability increase in our assets. And reducing the operational cost can have a massive impact on the levelized cost of energy, and especially when we talk about offshore wind turbines, where the logistical implications of going and fixing a turbine are very big, then the logistical and the costs associated with it. So the earlier we know something happens, it's so much useful to operate our assets optimally. So, yeah, we're looking forward to actually testing and deploying these, and hopefully have value delivered. [Angela Belcher] That's great. So we have about one more minute, if there's any other closing statements.
It really sounds like this has been a very fruitful collaboration, where the expertise on both sides have come together to share data and share science and technology that can make a difference. Is there—coming together on this project, what do you think has been the most important combination? [Kalyan Veeramachaneni] I think the most important combination has been collaboration through open source software and through working with the team on almost regular basis, a biweekly basis. And I don't— I think that's the most important thing for us. [Sofia Koukoura] I agree with Kalyan, because academia gives you so much more freedom to explore ideas in a way that often, in industry, we don't have much time. So that collaboration between bridging what we need for our applications and people who have so much more experience in machine learning and innovating, yeah, I think bridging those two is really important.
And as someone who has worked in both fields, I can really see the value of it. [Angela Belcher] Well, thank you both very much for speaking to us today. And we all look forward to seeing more of that your future collaboration and your future progress. Thank you.
So I'm pleased to move on to the next speakers in the panel, Energy Systems, SESAME Holistic Modeling of Pathways to Net-Zero Carbon Systems. And we have three panelists today. We have Emre Gencer, Dan Cherney, and Jen Morris. Dan is the Section Head of Emerging Energy Sciences at Exxon-Mobil. Emre is a research scientist at MIT
Energy Initiative. And Jen Morris is a research scientist at MITEI, and MIT's joint program on Science and Policy of Global Change. So welcome. Let's jump right in. Emre, can you tell us about the problems you're trying to address and the SESAME tools? [Emre Gencer] Absolutely, thank you, Angela.
And so we are really in the midst of a huge change in the energy sector. And one of the challenges that we see today is the need to satisfy growing energy demands while simultaneously achieving significant reductions in greenhouse gas emissions. And these emissions are from production, delivery, and consumption of energy, so the entire value chain. And one important transformation that we are seeing right now is about the greater convergence of power sectors, transportation, industrial, and building sectors, and also intersectoral integration. And to achieve the level of decarbonization we need for mitigating climate change, actually, we really need to take a systems approach and look at the carbon footprint of the energy system as a whole. And to address this challenge, at MIT Energy Initiative, we have developed a novel software, SESAME, which is short for Sustainable Energy Systems Analysis Modeling Environment.
And this tool is an online tool developed to explore the impacts of relevant technological, operational, temporal, and regional characteristics of the evolving energy system. And with this platform, we are focusing on accurate estimation of lifecycle greenhouse gas emissions, technoeconomic assessment, as well as the scalability and feasibility of emerging technologies. And as we continue developing SESAME, we are, of course, using it to analyze many important and contemporary problems. And one recent example is the MIT Energy Initiative's Mobility of the Future study. Professor Armstrong mentioned about this study in his opening remarks.
And for this study, we have performed an environmental impact analysis of various vehicle power trains and fuel options, in collaboration with Joint Program and Jen. And one example was, for example, we try to understand the emissions reduction potential, or switching to EVs versus hybrid vehicles, and how these results vary across time and across different parts of the US. [Angela Belcher] OK, thank you. Jen, can you talk a little bit about your different angle on modeling energy transitions and how it fits into what we heard about SESAME? [Jen Morris] Yes, great, thanks, Angela.
So at the MIT Joint Program on the Science and Policy of Global Change, we develop and use a global energy economic model called EPPA, or the Economic Projection and Policy Analysis model. And so the EPPA model takes a big picture, more aggregated view of the energy transition challenge. It represents the world, the whole world in 18 regions, and covers all sectors of the economy with a focus on energy representing many alternative low carbon technologies. And so this model captures global markets, interactions between regions and sectors, including trade and price effects, as well as impacts on GDP. And it's designed to explore future projections of economies, energy, and emissions under different possible scenarios. And we do a lot of scenario development to explore different energy and climate policies.
And so this broader view, with an economic focus, is very complementary to the detailed assessment approach that Emre described. And so while SESAME can assess, in great detail, alternative technologies and their implications for emissions, EPPA focuses on costs and the competition of alternative technologies within a portfolio of options. And we really see great potential in linking these tools. And as Emre mentioned, we first did so within the Mobility of the Future study. In that study, projections from EPPA on emissions intensity of electricity generation under different policy scenarios were input into SESAME's detailed fleet models to assess emission impacts of electric vehicles.
And so we're currently working together to expand the linkages between our models in terms of the types of information that EPPA can pass to SESAME and ways that SESAME can inform the assumptions in EPPA. And we believe that these linkages can really provide a better insight into the energy transition and the potential for low carbon technologies at scale. [Angela Belcher] Thank you for telling us about that, and also about the integration.
Dan, I want to bring you in. So Exxon has sponsored SESAME and EPPA, as well as Mobility of the Future study. Can you tell us a little bit about your and Exxon's interest in this kind of work, and what you think the main benefit is from working with MIT researchers and particularly the ones that are here today. [Dan Cherney] Absolutely, thanks, Angela.
So first and foremost, MIT is a world leading institution with a focus on energy. And as you mentioned, Exxon is really interested in understanding the energy transition that's, obviously, underway. Categories like energy storage, future mobility, carbon capture, all of these are critical to understand how the world is going to evolve. And building a little bit on points that both Emre and Jen made, the answer, going forward, is not necessarily obvious. One of the studies that we've been able to accomplish with SESAME is looking at emissions of power sector as you add more renewables. And it's become very obvious that just adding renewables to the electricity grid doesn't solve the emissions problem by itself.
And it's really critical that energy storage help play a role, because turning fossil sources on and off is not the best thing for either efficiency or for emissions. So we've been just delighted to be able to work with the quality people at MIT to study the solutions for the future and help us to promote decarbonization in both an informed and rational way, using a very good, sound model that we can share with the world. [Angela Belcher] So it sounds like there's a good working relationship back and forth between needs in industry and integration and collaboration with the modeling and other perspectives from the MIT team. Emre and Jen, can you, in the last closing minutes, talk about the importance in the connection with industry and with Exxon, in this case.
[Emre Gencer] Absolutely, maybe I can start first. So I will summarize in just one minute. So basically, I think there are really three important pieces here. The first one is really understanding the relevant problem. So when you close to collaborate with industry, you can focus on important problems that is contemporary and that you need to develop a solution.
The second one is you really understand the challenges to implement a solution. So in academia, we can come up with solutions. But if they are not implementable, they will not be as valuable, especially when we are in a climate crisis. And finally, we learn from expertise from R&D in these companies. And in Exxon-Mobil, so we are working closely with our colleagues with similar backgrounds and from similar institutions.
And we learn from each other. And I think this interaction is really important. [Jen Morris] Yes, I would echo what Emre said. These connections with industrial sponsors and Exxon, in this case, are incredibly valuable. They provide reality checks, inputs into our technoeconomic assumptions, and as Emre said, insights into real world challenges and research topics that are particularly relevant for applications.
And so these connections and collaborations really help ensure that our work is relevant and can make an impact. [Angela Belcher] Well, thank you both for— thank you, all three of you, for your exciting research and in sharing the importance of the collaboration today. We have one more panel in this section. And that's on Nuclear Power, Solving Operational and Maintenance Problems for the 21st Century Light-Water Reactors. And I'm pleased to introduce our two panelists for this last topic, it's Jake Jurewicz and he's managing corporate strategy of Exelon, and Michael Short, associate professor at MIT Nuclear Science and Engineering. So welcome to both of you.
And thank you for joining us. I wanted to start by asking what the challenges are in materials and longevity and what brought you two together? Michael, actually, you take that one. [Michael Short] Sure, yeah, so there are long standing decadal of multidecadal challenges in materials reliability for nuclear systems. The one that brought Jake and I together, we affectionately and professionally termed CRUD.
It's the deposition of corrosion products from elsewhere in the reactor on the nuclear fuel. In essence, things stick where you don't want them to. And that can cause accelerated corrosion. It can cause problems with power management of the reactor.
And ultimately, it leads to more downtime, less capacity factor. And what brought me into this, actually, was this problem has been looked at for 50 years, since its inception. And it had been largely empirical approach taken by industry. But when Exelon stepped in, they wanted to take a more science-based approach to saying, you know what? This problem is still here. What can we do about it? So Jake, I hand it off to you to fill in from Exelon. [Jake Jurewicz] And I think what originally brought Exelon to MIT and the low carbon energy centers was this desire at the executive level, and throughout the company, to engage in technology development at a much earlier stage than we previously had and most US utilities, historically, have.
The motivation for this project in particular is Exelon's known for, we have a large nuclear fleet operating in the Midwest and mid-Atlantic. And a lot of our plants have been challenged economically just based on market design issues and how we compensate— how we, oftentimes, fail to compensate zero carbon generators in the United States. So it was really very much kind of an all hands on deck approach to how can we keep these plants alive. How could take a more aggressive approach to finding some solutions to help reduce O&M costs to the plant? And that's what brought us to this specific project, amongst many others.
[Angela Belcher] Thank you. So Mike, when I think about work in academia, I think about kind of smallish scales materials or reactions that you're making at a benchtop. And I also think about how do you balance more of the basic science for the best way to work with a PhD student, and the need, in this case, to solve a really important problem on an industrial scale. So what do you think about when you're balancing those ideas and working with graduate students? Do you see it as a— do you see major benefits to a graduate student for seeing the full spectrum from the basic science to integration into an industrial problem? [Michael Short] Absolutely, I totally agree that we're— as a group leader, I'm always trying to balance inquiry and impact for my students. Inquiry, so that they can pursue lofty scientific goals, discover the undiscovered.
Impact, so they know that their work actually means something to a lot of people, and that they're helping society. And it's difficult to find the right combination of project and sponsor to strike that balance and actually make the student really feel like, wow, I'm contributing both to the field in general and to solving a specific problem. And that's what's been great about this Exelon project is we had that we had to get down to the nitty gritty about what makes things stick in nuclear reactors. We were doing some things, like you mentioned, making tiny samples of new materials in the lab, testing their stickiness on the atomic scale.
But we also built a reactor without the radiation. We call it the CRUD loop. It's specifically for testing and integrated effects as similar to reactor as possible without the radiation. Will these coatings survive in the high temperature, high pressure, aggressive chemistry, fast fluid flow, and boiling heat flux. And for a student to be able to see everything through from a technology readiness level of negative 7, when you're just making tiny little things in the lab, all the way up to the handoff to industry, it's immensely gratifying, both for the students and for myself.
[Angela Belcher] Thank you. And Jake, you've had some experience with MIT. Was that helpful in making these collaborations, coming back to MIT, and knowing from what you've done and your experience at MIT, was it easier in making these connections between academia and translation? [Jake Jurewicz] I think it certainly helped expedite some of the cultural learning that had to happen on both sides of the house, particularly for Exelon.
It was easier to help— it was easier to help navigate through here are the kinds of projects, the kinds of PIs, the kind of relationship that Exelon should try to strike with MIT. And it is easier to know knowing Mike short and a lot of the other MIT professors across the Institute and across MITEI, it was easier to strike up fruitful conversations and project opportunities. But I think, in particular, it made it easier to relate to the grad students, and the PhD students, and the research scientists who are working on the project directly, but underneath the professors.
Having been in that role before, I think it very quickly helped to resonate with them, because they got to see directly the connection to industry as opposed to just putting out another academic paper. The interaction was much more consistent, much more fluid for seeing how the work they're doing in the lab, which, sometimes, you kind of lose track of reality when you're a student in the lab, where it's actually applicable in industry and how valuable the work, and all the hours in the lab really can be. So I think it was really strong motivating factor for everyone. [Angela Belcher] That sounds great. I definitely know with my own students, when they can see an application that can make an impact on the world on a short timescale, even a long timescale, it really drives the excitement and the motivation for the project.
Is this also— Mike, is this also a good recruitment tool? Or is this also a good way of helping students think through and navigate future careers, thinking about academia and thinking about industry, and maybe even a connection between them? [Michael Short] Absolutely, because I think there ought to be more connections between them. That's the whole idea of this session. And allowing students to have frequent primary contact with an industrial sponsor, see that not only are there— that they are not just first authors on the papers, but on the patents that we're racing towards commercialization for these things.
It's an enormous motivator. It gives the students a lot of latitude to explore potential career paths, who they want to work with, what they want to do with their lives. I mean, I'd like to think that we can— I would like to think, I don't know, that we could solve any problem that we set our minds to. But the real question is, what do you choose to set your minds to? And seeing problems like this, where you can see, if I solve this scientific issue here, then there's someone waiting immediately to snap it up and make more carbon-free power. I can't think of a bigger motivator for someone who wants to solve the climate change problem. And I see that with our students, that they're lead authors on the papers, that lead authors on the patents, they're front and center whenever we talk about these problems.
It's great. [Angela Belcher] Well, we have just 2 minutes left on this. So maybe this question needs more than two minutes. But it's pretty fascinating that the two of you coming together were looking at a problem that existed for 40 or 50 years and have made an immense amount of progress on it over a short period of time. Do you think that is from the collaboration? Or is it also a combination of the collaboration and in the timing is right? I am interested in what you're going to say to this, Mike. [Michael Short] I would say it's kind of a combination of factors.
It's always a bit of a combination of timing, motivation, inspiration, you're always standing on the shoulders of giants of immense amount of research that's been done in decades past. This specific project kind of came about originally through another professor, Jacopo Buongiorno, who was submitting an ARPA-E proposal that we got an early look at, that was built on a lot of background research that Mike, and Jacopo, and others had been doing with other industrial sponsors. And it, through some back and forth and iteration, and we kind of massaged the scope and the key factors a little bit, we landed on the project that we did. [Jake Jurewicz] And Mike, I know you took some interesting inspiration, also globally, as you progressed in research and development.
[Michael Short] Absolutely, and I'd say, it's not like industry hadn't been tackling this problem. I mean, a lot of smart people had spent a lot of time trying to stop the formation of CRUD. But what you do in industry, when some of the previous panel said, you don't have time to delve into the sciences. You try the quick solutions first.
And if those don't work, you try the next less quick solution. And in this case, we continued and continued and continued. Eventually, we just had to ask the most fundamental question, what makes things sticky in nuclear reactors.
And at that point, you delve well into the realm of fundamental science. I mean, the nucleus of this project was one of my students and I staring at an equation that was 50 years old saying, let's exploit that difference of variables in a numerator. It actually came down to exploitation of a pair of variables.
But in the end, once we realized, OK, now we've got an angle to be constantly pulled and refocused by Exelon to focus, eyes on the prize, the goal is to get a technology out there, the papers naturally evolved from that. But it was so great to sort of have always the refocusing from industry to make impact. That was critical throughout the whole way. And honestly, it helped speed the progress of the research as well. [Angela Belcher] Well, thank you. So I actually have a lot of hope after hearing all three of those examples of really powerful collaborations with the whole idea of wanting to make an impact on the world and how can we work together to have— to do that in the most efficient and fastest way.
So I'd like to thank everyone for contributing today. And most of all, I'd like to thank you for your research on such important topics. So if you want to put some questions into the chat, you can go ahead and type those in now. And I'll start with a question about SESAME. And the question is, how does one include all costs of offshore wind, including connections to the grid, in comparing with building solar, and storage, and additional benefits? [Emre Gencer] So when we compare any alternative, we are really including all the costs. So if—I can start with an example from emissions.
So when you compare EVs with hybrids, for example, you cannot only look into tailpipe emissions. You really need to understand how you produce electricity, what the total emissions, all the way to driving that car. And similarly, when you do a cost analysis, you need to include all the components that goes in. So if it's an offshore wind, then you need to take into account the distance from the shore, and what will be the transmission line cost, so on, and so forth.
So basically, the type of analysis we are doing captures all the components in terms of cost and emissions to be able to have apples to apples comparison. [Dan Cherney] Yeah, and there's also a visual output, as Emre was saying, of all of the variables. And so the person that's asking the question and wants to make a comparison can see the impact of the choices that are made by the individual. So you can see what really matters for the cost or the emissions compared to the other variables. [Angela Belcher] Thank you.
Another question is about data machine learning comparison between— that was figured out, comparing that with learning from an experienced person directly. [Kalyan Veeramachaneni] I can answer that question a little bit. I think when we process the data sources, historically, what we have been doing is just automating that process, trying to do as much automation, maybe this signal in this frequency band is important, and that's how we should do it, process it, and use it for machine learning. And it was very accuracy driven, how accurate is a predictive model. What we are realizing is that the experts, domain experts, have that knowledge. They know when it's a vibration sensor and a particular signal, what sort of band we should be watching for an anomaly.
So we are now collecting that kind of input through our software interfaces from the experts. That has been the driver. And that has resulted in a significant change in how we approach machine for these problems. [Angela Belcher] Thank you. Does anyone else have one to chime— [Emre Gencer] Yes, so I think this is a very important question, not only for machine learning, but, in general, for modeling.
Because modeling, if you don't understand the physics, you can build models that will generate some results. But how to interpret these results will be the key here. And if you know about the physics, either for the applications that Kalyan mentioned, or for applications that we are tackling in SESAME, actually, you can really improve these models. And you can specifically find datasets and find relationships that will help you get the insights that you are interested in. So that's also what exactly we are doing.
So we are—in SESAME modeling, we are using process simulations, real world data, and so forth, to enhance our analysis and our modeling expertise. [Angela Belcher] Thank you. The next question, I think, would be for Jen. What policy was the worst among the EPPA scenarios? [Jen Morris] Well, we run a wide range of policies. And the worst tends to be the no policy, business as usual scenario, where there is very limited action taken in terms of emission mitigation or policies for low carbon technologies.
And as expected, that type of scenario results in a world that continues use of fossil energy sources. And then you can see how that translates into emissions and emissions intensity. And you can look at how that would impact the grid and the use of electric vehicles. So if you're pushing electric vehicles, but you're not also decarbonizing the electricity grid, then you're not really solving the problem if your electric vehicles are powered by coal generation. So that's an example. But yeah, we explore a wide range of scenarios.
And of course, those that involve policy that reduce emissions fare better. [Angela Belcher] Thank you. There's another question around SESAME. And the person said their aim is to shape technology, clean energy, and storage, so they could optimally contribute to global decarbonization. Are SESAME and other simulators available to outsiders to gain feedback about evolving designs? [Emre Gencer] Yes, with SESAME tool, we are trying to build an open access platform. So it will be open to the public.
And also it will be open to decision makers, policymakers, to be able to exactly answer these questions. So we are currently working hard to make sure that it is bulletproof and it's ready for public use. We are almost there, but not quite.
[Jen Morris] We do also have a publicly available version of the EPPA model that some people have been ambitious enough to take on and try to run and understand. An example is the Bank of Canada. And we usually have some interactions with them.
And it can evolve into collaborations that can be quite fruitful. [Angela Belcher] One more question about SESAME and EPPA is, can the two of you say something about comparison to other models like MARKAL/TIMES, NEMS? Can you do a comparison to some of the other models that exist for simulation? [Emre Gencer] Dan, do you want to start, because I think the direct comparison is with EPPA. Then I can compliment. [Dan Cherney] Yeah, I guess I would say what we were looking to do with SESAME was to find the spot where we thought that there was a gap.
And that was having a model that combines emissions of the energy system, all of the energy system, or as much of it as we could get into it, along with the technoeconomic assessment of those technologies. So that we could really understand how to most efficiently and effectively decarbonize the system. And so there's no other model out there right now that does what we want SESAME to do.
So Emre, if you want to take the comparison with the NEMS, or ReEEDS or MARKAL, or any other models, go ahead. [Emre Gencer] So the main difference is we have a very, very high-resolution technology presence and technology representation, which is one of the key areas that we see as the solution for climate change mitigation. Because we need technology. And MIT is very strong in technology development.
So we wanted to capture that in a very detailed manner. But, of course, SESAME is not the only model that can solve all the problems or can be used to make all the decisions. So that's why we are working with EPPA model and with more scenario analysis tools that takes longer time, horizons, and also do projections for different policy measures. And here, I don't think there is, as Dan mentioned, there is not a single tool that is exactly doing what SESAME is doing. But SESAME is designed to be a complementary tool. Wherever we see a gap, we want to make sure that we can complement by SESAME.
But Jen, I think you should chime in here. [Jen Morris] Yeah, so on our end, compared to some of those other models you mentioned, like MARKAL and ReEEDS, those are energy sector models that focus on a particular piece of the economy, whether it's the electric power sector, or more broad energy sector. And so EPPA is really an economy-wide model, where we're capturing all of the sectors of the economy, as well as the globe and multiple regions, and so interactions between sectors and regions, which gives kind of the full, broader picture and how electricity interacts with transportation or industry, and the interactions in terms of the impacts on the prices of fuels, of electricity, and other things, including even agriculture, and more broad than a lot of these energy specific models. So it's kind of this bigger picture that allows a broader view of how economies might progress over time. So it's really well suited for these long range scenarios that can be fed into a lot of these other types of models, like SESAME. We also do connect with models like ReEDS.
But it's the bigger picture view that's capturing the global picture and key interactions between sectors. [Angela Belcher] Thank you. We have more questions on SESAME. We might come back to it. But I want to switch to nuclear for a minute.
And so this might go to Mike and Jake. It says, regarding the longevity of nuclear plants, what other areas is MIT investigating regarding neutron embrittlement, concrete, table conditions, monitoring, and containment. [Michael Short] You just named quite a few of them. For example, other aspects of the Exelon project were preventing the ingress of hydrogen into nuclear fuel to maintain its longevity, as well as increasing what's called critical heat flux, or the total amount of heat that the nuclear fuel can output before you reach problems with heat removal due to film boiling. We also have some neat machine learning thrusts. We've got—one of our Co-PIs, Matteo Bucci, has been looking at how can you use readings from elsewhere in a plant, let's say, outside of the core, to infer what's going on inside the core.
So like other panelists were saying about you're collecting terabytes of data on these wind turbines, what can we do with it? Same thing, nuclear plants are instrumented out the wazoo. And there is probably a lot of information to be gained from a flow sensor on one side that can tell you about the temperature on another side. It's just a matter of discovering those correlations. So we have everything from high fidelity neutron modeling, to materials reliability, to new models of regulation for fission, and soon, fusion power plants. And the list goes on.
[Angela Belcher] Thank you. So this is a question that can be for anyone on the panel. It's industry versus academia, where were the most impactful research emerge in the next decade? And so the answer might be both. But does anyone have any comments on that? [INTERPOSING VOICES] [Dan Cherney] Go ahead. Go ahead, Mike. [Michael Short] OK, yeah, I could start.
I'd say using the CRUD problem as an archetype for the best problem to solve, I think industry with academia. with the industry's help, academia can help identify problems that have been lingering for decades, that have been constant roadblocks for which, instead of solving them, mitigation efforts are underdone or undertaken. These seem to be lower hanging fruit, something where the nature of research in academia can make a huge impact an industry very quickly. [Dan Cherney] I was actually going to say something relatively similar. I think that having academia and industry work together could be critically important for a couple of reasons.
Industry, people in industry are very familiar with the problems of operating some real world system. And oftentimes, people have been in that industry for decades and have lost their cold eyes at that looking at a problem for so long. And where academia comes in, you have all kinds of energetic people that are very intelligent, that are really eager to solve really important problems, and they bring a cold eyes, and an active creativity, and curiosity to solving real world problems that's just critical. [Emre Gencer] And I think we are really in— we don't have enough time to actually solve this problem. I think we are really in a hurry to solve the climate change issue.
And so everybody should work together. And I think there are brilliant people everywhere. So we can only tackle this problem if we work together. That brings up the next question, a similar question. And so it looks like you were getting ready to speak, Jake, so I'll let you have the first chance to answering this one.
It says, do you think the acceleration of the timeline to get to net-zero, do you think it has an impact on your collaboration in your engagement styles? [Jake Jurewicz] I would say, definitely, I would say I hope so. And I was going actually to say something a little bit, not necessarily controversial, but a little bit less like, oh, kumbaya, we should all work together. We should all work together.
But I would actually add from my experience in engaging in industrial sponsored research with MIT for the last five, six years, as well as other academic institutions, labs, startups, other big industry vendors, I think there is a lot more that could be done on both sides to really expedite the level of collaboration that's going on. At the end of the day, it really is going to be industry that will bring a lot of these new technologies to market, both incumbents and startups. And I think it's critical of our academic institutions, like MIT, given the strict timeline that we're on, to decarbonize the global economy, to really find ways of expediting paths to market.
I think we need to kind of redefine some of our KPIs and our definition of success of what is the purpose of an academic institution like MIT. MIT is really born and bred in industry-motivated, industry sponsored research and development. But I think there's more that can be done just given the evolution of how the venture capital system and the startup community and the innovation ecosystem has really evolved and developed over the last 20 years. And there's a lot of big guns that are starting to get pointed towards climate problems and climate tech just in the last six months.
I'm sure a lot of the folks on this call and in this panel have been following some of the activity that's been happening. So I would just say while I think it's been good, and we're making progress, I think we really need to expedite that progress. [Angela Belcher] Well, that was a fantastic summary. Does any of our academic collaborators want to take on that question, in terms of the accelerated timeline, and how that might have impacted the way that we collaborate or work together? [Michael Short] I could, sure.
So yeah, I'd say from the other side of the fence, I 100% agree with what Jake says. I think, ultimately, the mission of MIT is for the betterment of society. Everything we do is to make people's lives better. It's not—there is for the love of science, but again, it's what you choose to do so.
We at MIT produce products, models, technologies. And we try and push them into the real world. But as anyone who knows who's tried, you can't push a string. It's somebody has to be there pulling it. And that's where industry comes in is they know the problems, they know the markets.
But there also has to be something to pull on the other end of the string. And so making more of these connections in a concerted manner, with new funding models, new public-private partnerships, new industry-academia partnerships is critical to making impact and keeping Boston out of the water by 2050. [Angela Belcher] Well, I think we're ready to wrap up the session. It was an amazing panel. Thank you so much for contributing and the collaboration together. And I didn't get to take everyone's questions.
But there were more questions around SESAME. And maybe we can get those questions answered online. So I'd like to give a big round of applause to my panel. And with that, I'd like to turn it over to the next session. The moderator is Professor Anne White, professor and head of MIT Nuclear Science and Engineering, also a fantastic scholar, and leader, and a great partner at MIT. So Anne.
[Anne White] Thank you so much, Angie, for the very warm introduction. Good morning and good afternoon, everyone. This very next session is exciting. It is about cooperating on research for next generation technologies.
It's a pleasure to be moderating two panels in this session. One on solar energy research and one on fusion energy collaborations. There will be 20 minutes each for the panels, followed by a 20 minute combined Q&A session. We will end our session very promptly at 12:55 for closing remarks by Professor Maria Zuber. Our first panel, on solar energy research, features the speakers Miles Barr co-founder and Chief Technology Officer of Ubiquitous Energy, Vladimir Bulovic, the Fariborz Maseeh Chair in Emerging Technology, professor at MIT Electrical Engineering and Computer Science, Joel Jean, co-founder Swift Solar, Massimiliano, Max, Pieri, Cleantech Director, Eni Next, Tim Swagger, John D. MacArthur Professor of Chemistry at MIT.
Welcome, everyone. And thank you for being here today. Max, let's start with you. The Solar Frontier Center, or FSC, is one of the pillars of any cooperation with MIT.
And the SFC dates back to 2008. Could you tell us a little bit about how SFC research has evolved over time and some recent progress? [Max Pieri] Yeah, sure, thank you very much for that introduction. And indeed, the solar frontier center is one of the pillars of the cooperation with MIT. And it dates back to the very beginning of our cooperation with MIT.
We started SFC because we wanted to tackle a wide range of opportunities. And we felt that we needed to team up with a leading research institution like MIT. Up to now, we have carried out around the 25 projects within the solar frontier center.
And I would say that the research with SFC has significantly evolved over time, both in the scope and also in the nature of the projects. Initially, the scope was very, very broad, explore everything beyond silicone, polymers, quantum dots, solar concentrators, parabolic concentrators. Over time, we decided to narrow down the focus of our research.
And currently, we are focusing on a few technological areas, perovskites and polymers. But also, the nature of the projects that we are carrying out within SFC has changed. At the beginning, we focused more on early stage research.
And now we are going into more and more into technology deployment. A great example is for the World Bank flagship project that we are doing with Vladimir and Jeremy, which is about developing scalable printing and coating processes and materials for lightweight and flexible photovoltaics. So I would say a very important part of our cooperation with MITEI, both in terms of financial commitment, and also in terms of the human resources, the work that we put into this effort [Anne White] Thanks very much, Max.
Vladimir, let me come to you. Could you please tell us a little bit about how this partnership has benefited both Eni and MIT? [Vladimir Bulovic] Oh, absolutely. We have, at MIT, dramatically benefited from the relationship. There are over 500 student postdoc years, provided for by Eni during the lifetime of the existence of the solar frontier center. There are hundreds of research papers that came out of this.
And there are insights that our community has gained in that process that led us to start startup companies, Ubiquitous Energy and Swift Solar, led by Miles Barr and Joel Jean are just two examples. I'm delighted they are able to join us today. So they can tell you even more about their particular ways of learning through the solar frontier center, and then taking that knowledge out. Solar frontier center also informed us, in 2000 and— well, just a few years ago, let's say, when we published the Future of Solar study, where we looked at our knowledge base gained through the solar frontier center.
And we were actually generating that solar study at a cusp of the solar development. We were able, though, through the understanding of what we learned through the solar frontier center, and other knowledge around it, where the solar is going. And we're able to actually capture that in the study that still is relevant today. I would say, in so many ways, that one thing that I would emphasize is the commitment and steadfast commitment of partnership between Eni and MIT. It is that long term relationship that allows us to understand each other's needs, and allow us to evolve, at MIT, our focus. So to be more relevant for the Eni's final goals, and on our end, be able to adjust our fundamental research versus more applied research ratio, hence leading to the present state of the project, which is very much looking forward to the ways of scaling up ideas beyond just the labs.
[Anne White] Thanks very much for that. As Vladimir mentioned, Joel and Miles, you're both former MIT PhD students at the solar frontier center. And you've each launched spin out companies.
Joel, you founded Swift Solar, Miles, Ubiquitous Energy. Can you let us know a little bit more about that process and how being a part of MITEI played a role? Joel, how about let's start with you. And then we'll go to Miles. [Joel Jean] Yeah sure, so at Swift Solar, we're working on commercializing efficient and lightweight solar based on perovskite materials.
So I guess you're asking about connecting the dots looking backward. So I think a few experiences stand out from MIT. I did my PhD work on new solar technologies, quantum dots, to organics, and perovskites. And I think that technical experience supported my work a Swift, even things as simple as knowing how to maintain a glovebox or what equipment vendors are good to work with.
I mean Swift isn't directly based on technologies that we developed in the solar frontier center. But its seeds can kind of be traced back to, partly, to an Eni project where we made these super thin and lightweight solar cells and put them on a soap bubble, which is kind of useless. But it really did capture people's imagination and made me think deeply about lightweight solar and what it would take to actually make it real. Actually, I also spent a couple of years at MITEI. We're writing this Future of Solar study with Vladimir and many others. And then after my PhD, helped launch and recruit team members for this Tata-MIT GridEge Solar program, which was really focused on solar applications in developing countries, in India.
It was kind of a quasi-startup with MIT. And I think those two experiences really give me a chance to learn about some of the big picture questions around solar economics and policy, and really learn how to build a team and manage, even manage finances. I think, looking back, that's actually those things have been more important than my lab work for charting Swift strategy. And I think one final dot there is actually at MIT, I was part of a climate action group called fossil-free at MIT. And we actually were pushing back against fossil fuel funded disinformation around climate change.
And as much as that might sound anti-Eni, I think that's actually aligned with MIT and Eni's goal of helping fix climate. So Eni's supporting next-gen solar and fusion work, because they think it's important for climate. Obviously, that's helped MIT spin out some great startups.
But in the end, I think it still comes down to making these big capital investments in scaling and deploying clean energy. So I just hope they keep moving in the right direction. [Anne White] Thanks for that. Miles, let's come to you, maybe a little bit about the process of spinning out the company and connections where you feel MITEI played a role. [Miles Barr] Yeah, sure.
So Ubiquitous Energy, as our name implies, our goal is to put solar everywhere. We want to find ways of deploying solar on everyday products and services all around us. And the origins of that actually started during my PhD at MIT.
And I was really fortunate to actually be one of those first projects with the Eni-MIT solar frontier center back in 2008. At the time, we developed a way of putting solar cells on any surface. So we started out by demonstrating that on paper. We literally made a solar cell on a sheet of paper.
And we unveiled that, actually, in 2010 right around, I think, there was a public dedication of the solar frontier center. And the research got a lot of attention. And that was really exciting. That, I would say, kind of kicked off my own career in solar energy.
Through the solar frontier center, I had the opportunity to work pretty closely with industry, with Eni. We had researchers come into our lab and our team. And we made solar cells in our lab. I had the opportunity to go to Novara, their research center, test solar cells, work with their researchers to optimize our technology. And also I was fortunate to participate in some of the steering committee meetings with the PIs at Eni. And so all of that experience, I think, was really valuable for me personally in seeing how the industry was kind of informing where the technology should go.
And so really setting metrics that were relevant for commercialization, relevant for deployment, we went and we demonstrated these cells to the leadership of ENI, the leadership of MIT, and that was really exciting. All that work inspired me to start a career to commercialize new solar technologies. And I was really inspired by this theme of solar everywhere.
And that was what was kind of the origins of Ubiquitous Energy. The MIT ecosystem really supported that. Learning how to start a company, how to raise money for a company, build a business plan, and then the technical training and the kind of industry lens that the Eni solar frontier center research group really helped craft that in a way that was going to be commercially viable. And so we're still at it today. We are making transparent solar today.
These are solar cells you can see through. You can put them on window glass, you can put them on consumer electronics. They generate electricity from the sun, just like normal solar. But you don't see them.
And so in our view, this allows us to open up lots more surface area for solar energy generation. And this really all started back in 2008, when the solar frontier center started. [Anne White] Thanks so much, Joel and Miles. Tim, let me come to you and ask a bit of a big picture question. How can these industrial points of view and this engagement really enrich research at a university? [Tim Swagger] Yeah, well, in chemistry, we're in a little basic science cocoon.
And it's always important that you reach out and see the greater world. And I think MIT has a history of bringing companies in. And certainly, MITEI has been a great vehicle for that, enlightened companies that tell us about the future, and what they see, and what they can do. So helping scientists to identify the future and where they can make the most impact is really important.
I would say, in the case of Eni, they've brought this to a whole new level of engagement with us and collaboration, basically through things that Vladimir emphasized, a long term relationship, getting to know a personal relationship, where you go back and forth and meet each other. And that's been just incredibly valuable. And the framework of the solar frontier center has really helped to steward that really, really nicely. And so a couple places where, I would say, there have been really profound impact for me, one has been in the organic photovoltaic area, where people are making complicated polymers. There are many efforts in China, where they're making insanely complicated polymers that are not realistic to be commercialized.
I mean, they're setting records. But in talking with any scientist, in particular Riccardo Po, he had, essentially, come up with ways to quantitate out what he called synthetic complexity, where you kind of do trade-offs between efficiency, and the complexity of how you make things, and the costs associated with it. He came up with a very fundamental nice rubric on how to actually quantify this. And I was so impressed I was an editor for journal Macromolecules at the time. And I convinced him to write a prospective article to get that out to the entire world.
And so that was a place where this collaboration actually went out to do good for the planet. And that's a very highly cited article. It's something that caused us to rethink how we do synthesis, and we come up with some new reactions and new materials.
And the fact is, through the flagship program right now with Eni, we're evaluating [INAUDIBLE].. Another place where they've really helped, Eni has started, independent of MIT, an effort on solar concentrators. And they wanted to look more broadly. They actually have some outstanding technology in that. And we make polymers that are used in sensors.
One of the ways we do that is to essentially capture light and transport the energy around to create signal gains. And the same method came to us, we can use in some solar concentrators. So again, they helped us to kind of see a application of some physics in a new area that we hadn't seen before.
This collaboration has really been inspiring. And through there are other— others that many of my colleagues could talk about. But that's the way I've seen it really impact basic science. [Vladimir Bulovic] Anne, if I can add to Tim's works. I have known Tim now for 20 years.
And I'm an electrical engineer. Why would an electrical engineer spend time with a chemist? Well, it's programs like this. It is programs like the Eni program that thrust us upon each other and make us realize that combined knowledge can lead us to a higher level of appreciation of what technology can deliver than what we can do alone. So again, I commend Eni for giving us a chance to build these partnerships throughout MIT. There are on the order of 25, 30 professors that participated in the span of the program in the program. And they span, if I'm going to make a guess, about 10 different departments.
So that is what the program allowed us to do. We even have pooled the resources, in the very beginning of the program, to build a central facility that is the shared experimental facility, initially in building 13, now in building 12, that is providing for the nanomaterials development and solar testing. Anyone can use it. And indeed, hundreds of people, not just in the program, but across MIT, that use the facility to get their work done. Again, thanks to Eni. So I'm tremendously grateful for the chance that we had a chance to encounter each other back in 2007, the year before we started the program, in March 2008.
And every year, we go to Novara or Milano and go ahead and spend time asking questions of our colleagues. We spend about three days every year and try to identify what are the next set of opportunities. Those are incredibly valuable times to me. Having a chance to welcome them back to MIT, that happens throughout the year as well. Max is a permanent fixture at MIT.
He actually lives here. He's a permanent conduit to Eni. But beyond him, many of his colleagues do stop by. Some of them spend time in our labs, synthesizing new chemicals. Some of them come by to learn about the latest thing we've done and to give us advice on the things that they would care about.
And to reform their own thinking, and our own thinking. So that's what a really good relationship is about, recognizing the value of each other as equals. And then recognizing the ways to connect at a different level that none of us alone would be able to deliver on our ow