Phil De Luna
National Research Council Canada
Reaching Net Zero with AI and Robotics
Reaching Net Zero with AI and Robotics
Reaching Net Zero with AI and Robotics
Is net zero by 2050 really possible? Phil De Luna, Director of the Materials for Clean Fuels Challenge Program within the National Research Council of Canada, lays out what the path to a decarbonized society looks like. Since the onset on the Industrial Revolution, upward trends of CO2 emissions and gross domestic product per capita have matched nearly 1:1. Cheap fossil fuels have allowed us to enjoy our current high standard quality of life. But if global temperature are to stabilize, we must learn how decouple carbon emissions and GDP, reshaping our economy so that carbon is not so pervasive. How can we continue to be productive as a society without emitting CO2?
- According to Phil, only 25-30% of all carbon emissions come from electricity generation. The rest results from a myriad of other sources, including industry like steel and concrete, transportation, land use changes like deforestation, agriculture, and building heating. Phil lays out 5 key steps to holistically reduce emissions – conserve natural carbon sinks, increase renewable capacity, electrify as much as possible, focus on new methods for hard-to-abate sectors like steel and cement manufacture, and carbon capture technologies.
- As stated in the 2020 Energy Technologies Perspectives report from the International Energy Agency, almost half of the emissions reductions pathways required to get to net zero rely on technologies that are not yet commercial. Phil is adamant that we must speed up this discovery life cycle if we are to successfully decarbonize by 2050. Luckily, we live in an age of new advancements that can accelerate technology production – including increased computing power, a data-rich world, and robotics.
- At the National Research Council of Canada, Phil uses what he calls a materials acceleration platform to marry robotics, high throughput experimentation and AI to close the loop on innovation. By using a robust, automated process, his lab can accelerate the discovery of new materials and technologies to speed up the energy transition, removing human error from the equation.
Thank you very much. And thank you everyone for watching this talk today. Today I’m going to talk a little bit about is net zero by 2050. possible. And more specifically, how can we speed up discovery to get to net zero? What do we need to do to get there as fast as possible? So I’d like to begin this by prefacing this in the problem that we face. These are three graphs that I’m sure you all know very well, by now, global temperatures from 1850, to today, co2 emissions from 1850. Today, and gross domestic product per capita worldwide, from 1850. To around today, you’ll see that each of these three graphs match almost identically. And the fact of the matter is, we owe our quality of life, our standard of living our technology, and, and quite frankly, the the wealth that we have in our nations to the Industrial Revolution. And of course, the energy that was produced by cheap fossil fuels that have allowed us to create the society we live in today. The problem is, of course, that we were too good at harnessing this energy, and not properly looking into the effects and impacts of the emissions that came out of using fossil fuels. The big challenge that we have today is decoupling gross domestic product, and co2 emissions so that our global temperatures can stabilize. Gross Domestic Product and co2 emissions have been correlated almost one to one since the Industrial Revolution. The only way to break these two things apart, is to start thinking about ways to be productive as a society without emitting co2, of creating energy creating products, creating materials that also are sustainable or do not in themselves, emit carbon emissions. This is a tremendously difficult task because carbon emissions are embedded in every part of our society and every part of our economy. This is where co2 emissions come from today, about 20 to 30% 25 to 30% comes from electricity generation 13 Giga tons or 13 billion tons of co2. When people think of renewables, or when people think of energy, they think of solar and wind. But this is just part of the problem. The majority of co2 emissions actually come from non electricity generating sources like industry, which includes which includes steel and concrete making. And these other areas. Transportation, which is both lightweight transportation, freight, as well as air travel, land use change, which is the clearing of Amazon rain forests in order to make room for a ranching, for example, or other forms of land use changes, just forest fires due to increase in temperatures, or the spill on effects. Agriculture, which of course is due to, surprisingly enough, the amount of meat that we consume. Cows do a lot of burping and quite frankly, a lot of the methane that comes from cows and just their everyday living contributes to the co2 emissions or the carbon emissions, greenhouse gases that we see. And then buildings, of course, which is to heat our buildings, especially in cold climate conditions, like in Canada, where I live. So the question that I’ve I’ve, that I asked, and that people have always asked me is can we really decarbonize our economy, especially when a large majority of our economy is reliant on these processes that emit co2, and that actually have nothing at all to do with electricity generation. co2 emissions are pervasive. So what can we do? Well, the first step is to protect what we have. Let’s try to conserve the natural carbon sinks, whether that’s Amazon rainforest, whether that’s wetlands, let’s try to conserve the land that we do have. The second step is to increase renewable capacity everywhere. This is already happening, solar and wind is already cheaper than many forms of coal around the world. In fact, no one could have predicted the decrease in prices for solar electricity over the last few decades. This is a tremendously exciting opportunity to transition, especially in new developing economies and energy systems that have not yet been established or invested in towards renewables. The third is to electrify as much as we can electrify everything, whether it’s using heat pumps in homes to electrify heating, or whether that’s electrifying our transportation and our lightweight mobility through electric vehicles.
Today, electric vehicle models are booming. Why? Last year, Norway actually purchased more electric vehicles than gas consuming vehicles. many jurisdictions and governments around the world are outright banning petroleum based or fossil fuel based combustion engines by 2035, or even earlier than that. It’s clear that to electrify things, we need to change the way that we move energy around our systems. Step four, of course, is to tackle hard to abate sectors. These are sectors where it is impossible or difficult to to address and decarbonize using electricity alone. Steel manufacturer, fertilizer manufacturer, cement manufacturer, what we call process emissions, or emissions that result from simply this the material being made, are very difficult to decarbonize. And therefore, we need new technologies to fundamentally change these sectors, technologies, which have yet to be invented. And then finally, we have to carbon capture carbon from the atmosphere. And we have to do this using forms of carbon capture, such as direct air capture, which is faster and uses a smaller footprint than nature based solutions. In 2020, the International Energy Agency published a report called energy technologies perspectives, where they looked at over 800 technology pathways to get from now to net zero, what they found was we’re going to need a varied mix of different solutions. And that there is not one single solution that will get us to net zero, rather a combination of solutions that make the most sense, depending on the jurisdiction, and the country and resources with which you live and have access to. But of course, we need transformative technology to get there. From that same report, it was determined that almost half of the emissions reductions that we need to get to 2050 rely on technologies that are not yet commercial, they do not yet exist. They’re either at the lab scale, the prototype or the demonstration. What’s more, is that in order to get to 2050, we are talking a massive infrastructure investment, we will need to build the world’s largest solar Park every two days to get to that zero by 2050, the world’s largest electrolyzer in operation every hour, and the world’s largest carbon capture facility every week. This is a massive, massive infrastructure investment. And along that technology development lifecycle, you really have five stages in the lab, at the prototype at the demonstration, early adoption and then finally maturation. But of course, this takes lots of time. What you’re seeing on the screen are a few different technologies, and the time it took to go from invention in the lab to impact deployment, commercialization and widespread use. The car was invented around 1890 and it took seven years for that to become widespread. The cathode ray TV also took 44 years from invention to impact. As time went on, and advancements in technology came to be, we found that the second iteration took a shorter amount of time. For example, the LCD TV only took 20 years, and it was a second generation of the cathode ray TV. However, this scaling with technology and scaling with time does not necessarily hold true for energy. Nuclear Power took about 40 years has almost half a lifetime to go from invention to impact. solar photovoltaics, 55 years wind 40 years, we do not have for decades to get to net zero, we only have three, we do not have half a year to go from a prototype or from the lab to impact we need to accelerate this discovery.
This is how discovery is done today to just find one new material, one new battery, one new solar cell when you catalyst, you typically have a plan. You make the material, you process it, you characterize it, you test it and you analyze it. This cycle happens over and over again. And there’s no guarantee for success. a PhD student can spend an entire four years during this cycle just once. This is why it takes so long to go to from invention to impact for new clean energy technologies. And we have to do better. Thankfully, we live in an era where there are three underlying new advancements that are unlocking the way that we can discover. The first is that today we have the most computing power than ever before in history. The computing power in a cell phone is more than was the computing power used to bring man to the moon. Secondly, we live in a data rich world There is more data today than any time before in human history, data and computing are needed in order to fuel artificial intelligence, new algorithms and ways to leverage these technologies to predict new materials to predict new energies. And lastly, robotics have advanced to the point where not only are they more accessible, but they’re also cost effective and more and with enhanced performance. On the right you see spot, which is a robotic dog by Boston Dynamics. Today, you can go online and purchase a robotic dog and have it delivered to your doorstep. Never before in human history has this been possible. Because of this, we can now speed up this cycle by using robotics to help us with automating and accelerating the synthesis, the performance and the testing. And then we can use artificial intelligence to augment our data analysis in our prediction to help speed up this loop and make it more accessible and more faster to find new materials to find new technologies. And this is exactly what’s happening today. This is a video of what we’re calling a materials acceleration platform, or an automated robotic laboratory. This specific robotic laboratory was made by researchers at the University of British Columbia and the University of Toronto. At the National Research Council of Canada where I work, we will be working with these partners and more to develop a whole new suite of these materials, acceleration platforms, or baps. The concept here is to marry robotics, high throughput experimentation, and artificial intelligence. So we can close the loop on discovery and do things in an automated and more robust fashion. What you’re seeing here is a robot that is making a thin film material that can be used in either solar cells, photovoltaics, or light emitting diodes and next generation lighting. It’s synthesizing the material, mixing the precursors, hardening the film, and then testing the film for electricity conductance visible spectra and other important forms of physical characterization that we need to then correlate this data to predict the next material. What advantages we have here is not only are we able to do this on a more rapid and automated scale, but the data that we collect is consistent. It essentially removes human error from the equation. And more importantly, by accelerating this discovery, we can go from a factor of 1x 210 X, instead of spending $10,000,000.10 years to find the next material, we could spend $1 million in one year.
Unknown Speaker 12:52
We are really excited to be looking at this the future of accelerated materials discovery with these materials acceleration platforms, because for the first time in history, we have the computing power, we have the data, and we have the robotics to actually predict materials in a new way. So the question that I asked you once again, is net zero by 2050. possible, it has to be. And I end this presentation with views from three different perspectives, from finance, from oil and gas sector and from climate activists who are demanding that this transition occurred. Today we’re seeing the financial sector and private markets support clean energy technologies in a whole new way. Larry Fink the BlackRock CEO, the world’s largest asset manager, had recently said that the climate crisis will reshape finance, that their investment thesis will be centered around sustainability, knowing full well that the future will need to be sustainable. And this represents a tremendous opportunity. Because of this only gas executives, for example, the shell CEO recently said that for Shell Oil is past their peak, they will now transition like many other European oil and gas companies to becoming a whole broad based energy company, rather than just an oil and gas company. And finally, the youth of today are the ones who are demanding and pushing this because the youth are the ones who will be impacted by climate and more intensely as time goes on. Greta tunberg, for example, recently said I’m not here to make deals, I’m not a politician, I make no money. I’m here just to advocate for the transition to a clean energy economy to a low carbon future. And I truly believe that in order to do this, we need new technologies. These technologies have yet to be developed, and we don’t have the time that we traditionally would to develop them. So therefore we need to utilize data, robotics and artificial intelligence. Help us speed up these discoveries so that we can actually get to that netzero future by 2050. Thank you