Play Live Radio
Next Up:
0:00
0:00
Available On Air Stations
Watch Live

KPBS Midday Edition Segments

Bay Area Tech Nonprofit To Track Power Plant Pollution With Artificial Intelligence

 June 6, 2019 at 10:43 AM PDT

Speaker 1: 00:00 The announcement is startling air pollution, including carbon emissions from every power plant in the world will be precisely tracked in real time. This comes from a Bay area nonprofit called what time? With a one point $7 million grant from Google, what time plans to use satellite imagery and artificial intelligence to track plant volution and then make its data public as part of coverage from the KPBS climate change desk. Kyle Bornstein operations and partnership associate at what time? Spoke to KPBS is mark Sauer by Skype. Here's that interview. Speaker 2: 00:35 Well, let's start by describing how, what time we'll track greenhouse gas emissions, what high tech tools are used. Speaker 3: 00:42 So what we're going to be doing is actually leveraging existing satellite technology, uh, one of them being the Copernicus, which is the satellite that Europe launched. And we're going to be using a infrared sensors and image detection kind of coupled with artificial intelligence and machine learning technology to sort of measure based on the heat rate coming out of the actual smoke stacks of fossil fuel power plants, what the total emissions are of, of virtually every power plant in the world. Yeah. Speaker 2: 01:13 And why do you think a database like this as needed Speaker 3: 01:16 in the u s and Europe and maybe Australia, you have bodies that are regulating this, but in places like Russia, Saudi Arabia and India, person x on street, why doesn't really know what's causing the health hazards and the damage to their own life into their own ecosystems. So if we can open source that data, people will be held accountable and everyday citizens will really know what's going on Speaker 2: 01:40 and how reliable we'll the pollution data be. I imagine there will be a lot of skepticism that these precise readings and power plants can actually be done continuously. Speaker 3: 01:49 Yes, there, there there's, there's definitely gonna be, you know, a level of precision there that over time will become better and better. But, but we believe the, with the three different groups involved world resources, as you watch time and carbon tracker and sort of the NGO approach to this, the, the, the nonprofit approach to this, we're going to be as precise and as analytical and empirical as possible to really make sure that the data that's open sourced can be used in such realms. Like, I don't know, maybe the Paris Agreement, uh, to hold accountable people on their, their actual power plant emissions Speaker 2: 02:25 and what pushback can be expected from the power plant owners themselves. Speaker 3: 02:29 So there could be some pushback, but we've actually seen our plants responding fairly well where they're interested in this data. Now they could be interested in it for purely financial reasons, but some actually want to know, hey, how am I doing relative to my neighbors in an emissions, uh, perspective? Uh, so, so limited push back. But the pushback could be, well, hey, now I might not be able to make as much money if the power plant that operates a miles away from me can kind of know what's going on, uh, in, in, in my emissions because they can get a sense of, hey, at these moments, what am I bidding? Or what am I effectively selling power and the type of that power. Speaker 2: 03:11 Right. And uh, you would wonder, not only I'm worried about your competitors finding out data about you, but you'd also wonder if in certain places there might generate lawsuits saying, as you said, we're polluting this particular region. People are having asthma attacks, for example. And I'm suddenly their spotlight shined on you because they're looking at the pollution you're generating. Speaker 3: 03:32 Yeah, exactly. I mean you really hit it head on there. I'm mark because the people can effectively now hopefully have this reliable data set. And again, this is only one of the many different types of sectors, power plant emissions that are involved in such accountability in, in, in, in standards like the Paris agreement and other sort of environmental regulations. But really now plaintiffs in cases will effectively have that data. And yeah, there will be some scrutiny on how precise, but before this data is no longer, you know, wasn't available. And what's known now about the amount of pollution that comes from power plants around the world. And again, we're talking about fossil fuel plants in the United States for example, you have a, a body set up by the EPA. And this system, uh, is, is effectively sensors that are at the actual power plants, the fossil fuel power plants. And this is measuring sort of the total emissions coming out of sort of the plumes of smoke stacks. And they need to report this to the EPA every 30 to 60 days. So the equivalent of that also goes on in Australia, in Europe, but it doesn't go on everywhere else in the world. Now how big a problem Speaker 2: 04:43 do you, uh, determine it is that emissions are underreported or folks who are illegally polluting and certain places? How bad a problem is that? Speaker 3: 04:51 So, um, you know, my background is a environmental economics and policy and in global environmental governance and out energy and climate policy. So from, from the perspective of, of what I learned, uh, in, in, in what I think a lot of, you know, researchers and professors are preaching and regulators is that if we don't get warming under control, if we don't effectively curb any greenhouse gas emissions at an accelerated rate, climate change and global warming is here. It's now, but the actual, a magnitude of the potential harmful effects is what's coming. And if we don't act now, we can effectively have the ice caps melt at a pretty accelerated rate. The United States, Speaker 2: 05:34 as you mentioned, the EPA tracks is pretty well in other developed countries and countries that are allies and in compliance with this. But it might be startling for people to see in black and white just what this data shows as you start looking at every power plant everywhere. Speaker 3: 05:49 Exactly right. And if we can somehow put it, uh, you know, this mechanism in place where that data can be not only readily access understood by an everyday citizen, well then, hey, they can hold their, their, you know, stakeholders accountable. They are the rate payers ultimately paying to have energy in, in, you know, a great, wonderful life and, and utility to do. And basically the freedom to do anything they want. But if that's having negative consequences on their health and imposing a sort of a, of an appeasement on their lives, well, we'll then, hey, they should be able to do something about that. Speaker 2: 06:26 And B, need to know it to, uh, the, the first thing about getting a solution is to figure out the problem is Speaker 3: 06:31 yes, exactly. Speaker 2: 06:33 And finally, what's the timetable here? When will this data on power plant pollution begin getting posted? Speaker 3: 06:40 So I think we're about six to 12 months from sort of having a, an initial sort of way to measure that and to actually have sort of an idea. Let's say we pick one country first to really lock down the measurement that's needed for this. So the so-called equation if you will. And then eventually we're going to be rolling that out to the remainder of countries. I think maybe, you know, a year and a half to two. Uh, but the project is supposed to be funded for three years. But in terms of what's available now, they don't know what time we have this core sort of, uh, of technology around knowing when energy, every five minutes is better or worse for the environment. And that's available today. And, and, and consumers and corporates and regulators that we work with know about this. Um, and we're trying to make it more and more readily known to the public because they can actually do something about, hey, how clean is my energy right now? How much pollutants are in my, my uh, you know, if I flip a light switch in my home, what is the energy doing that I'm getting from the power grid? And how dirty is that? That was KPBS as Mark Sauer speaking with Kyle Bornstein operations in partnership associate. At what time.

Watt Time said it plans to use satellite imagery and artificial intelligence to track plant pollution and make its data public.
KPBS Midday Edition Segments