[00:00.000 --> 00:13.000]  Hello, you all, morning creatures. Good morning. We're about to start the talks of the day
[00:13.000 --> 00:19.000]  and I'll leave you with Tobias Augspurge, who will be talking about how open source
[00:19.000 --> 00:25.000]  can leverage more sustainability in the technology. So welcome, welcome.
[00:25.000 --> 00:33.000]  Yeah, welcome also from my side. So I'm actually working in atmospheric science.
[00:33.000 --> 00:41.000]  So by education, I'm an aerospace engineer. I'm in my day job calibrating satellite instruments
[00:41.000 --> 00:53.000]  for atmospheric science. But let's say in my private time, I do this professional hobby,
[00:53.000 --> 01:01.000]  so I'm actually investigating with some other people how does open source can have a significant impact
[01:01.000 --> 01:11.000]  on environmental sustainability. And we created this project within a small community called Prototypes
[01:11.000 --> 01:17.000]  that is actually investigating multiple ways how to increase the sustainability of our environment
[01:17.000 --> 01:25.000]  open source projects. And this project was started with a friend of mine and me about
[01:25.000 --> 01:31.000]  two and a half years ago within a pandemic. So we had a lot of time and we realized, okay,
[01:31.000 --> 01:39.000]  there's actually no overview about what open source really means in environmental sustainability.
[01:39.000 --> 01:46.000]  And so we started to create a huge list, an awesome list. I think you know what this is.
[01:46.000 --> 01:52.000]  That actually lists all projects that we can find in this domain. And quite soon,
[01:52.000 --> 01:59.000]  Irene joined this project because she did a similar development four years ago.
[01:59.000 --> 02:07.000]  And we were thinking about, okay, we created now this huge list and let's make a study about this.
[02:07.000 --> 02:15.000]  How can we quantify and how can we also find quantitative ways to investigate what actually
[02:15.000 --> 02:24.000]  open source means in environmental sustainability. So Irene was actually funded by SUBAC,
[02:24.000 --> 02:34.000]  that is an accelerator program to open up data in climate. And after that,
[02:34.000 --> 02:41.000]  Josh Hopkins from Open Corridor joined us. He is interested to integrate multiple of the projects
[02:41.000 --> 02:50.000]  we investigated in a larger software project also for environmental sustainability.
[02:50.000 --> 02:57.000]  So what does environmental sustainability in general means? It is actually the goal
[02:57.000 --> 03:05.000]  to preserve the natural resources that we all depend on for future generations.
[03:05.000 --> 03:12.000]  But what does it actually require? So it requires the intelligent intention to understand,
[03:12.000 --> 03:18.000]  predict and manage the stability of our natural resources. And that means that you need to
[03:18.000 --> 03:25.000]  understand that actually we depend on nature and the natural resources that have been created
[03:25.000 --> 03:36.000]  over a million of years by the nature itself and the natural earth systems that we see here.
[03:36.000 --> 03:43.000]  For example, you see here on the right hand side an atmospheric simulation by the open
[03:43.000 --> 03:49.000]  source project called SCREAM. They are investigating how does the content of water vapor
[03:49.000 --> 03:55.000]  change and how does water vapor actually change the climate in the future when our earth is
[03:55.000 --> 04:00.000]  heating up because when the earth is heating up, more water vapor will be in the atmosphere.
[04:00.000 --> 04:09.000]  And this is a huge problem because it actually accelerates climate itself because water vapor
[04:09.000 --> 04:16.000]  is itself a greenhouse gas. So we see there an actually accelerating process and software
[04:16.000 --> 04:26.000]  is there very important to actually predict and to understand how this will be in the future.
[04:26.000 --> 04:34.000]  So the core question of the study and analysis that we did is what enables open source software
[04:34.000 --> 04:40.000]  and environmental sustainability? So I think you know that open source is now dominating
[04:40.000 --> 04:47.000]  our software industries worldwide and there are studies from synopsis that shows that actually
[04:47.000 --> 04:54.000]  in 79% of the code bases worldwide you can find open source code and they did it because
[04:54.000 --> 05:00.000]  they do security auditing and they look into the code and they see okay we find it everywhere
[05:00.000 --> 05:07.000]  and it impacts all our software that we depend on worldwide. So what we did in our study is
[05:07.000 --> 05:15.000]  we started with the state. So what is actually the state of this ecosystem? And we created
[05:15.000 --> 05:22.000]  by multiple insights but we also wanted to derive some principles based on the multiple
[05:22.000 --> 05:29.000]  projects we investigated and interviews and people we talked to. But we also wanted to
[05:29.000 --> 05:34.000]  create some visions. So what are actually opportunities? What are strategies? What is the
[05:34.000 --> 05:46.000]  potential that we are missing that actually open source delivers us in environmental sustainability?
[05:46.000 --> 05:55.000]  So let's have a look about what are applications of environmental sustainability in open source
[05:55.000 --> 06:02.000]  and one very important aspect there is the environmental intelligence. So you get satellite
[06:02.000 --> 06:10.000]  data all around the world from multiple satellites, from NASA satellites, from ESA satellites
[06:10.000 --> 06:16.000]  and most of this especially satellite data that deals with the environment that is really
[06:16.000 --> 06:21.000]  looking at the environment is actually open to the public and you can process it. There
[06:21.000 --> 06:31.000]  is a lot of open source tools and with this multi-spectra and also with images and with
[06:31.000 --> 06:36.000]  today also spatial resolution and temporal resolution of those satellites you can derive
[06:36.000 --> 06:41.000]  multiple aspects of our environment like the air quality but also how our forest changed
[06:41.000 --> 06:50.000]  and also how noise actually is changing with our cities. You can measure but also and it's
[06:50.000 --> 06:55.000]  very important point source emissions that means who is actually creating emissions and
[06:55.000 --> 07:00.000]  you can directly point at this is the company who creates this amount of emissions and this
[07:00.000 --> 07:09.000]  is getting more and more an important topic because we have now multiple aspects that
[07:09.000 --> 07:16.000]  are coming like CO2, Texas and carbon trading where we actually need traceable data about
[07:16.000 --> 07:22.000]  what is the amount of emissions somebody created but on the other side there's also things
[07:22.000 --> 07:29.000]  like biodiversity that you can measure like satellites create images with not just RGB
[07:29.000 --> 07:34.000]  but they have a huge amount of different colors that they measure and if you combine those
[07:34.000 --> 07:42.000]  colors you can actually derive the change of biodiversity of forest or any living being
[07:42.000 --> 07:50.000]  on the planet. Another important topic is also the urban vegetation so how do we lose our
[07:50.000 --> 07:56.000]  urban vegetation and in combination with the heat islands because where we lose our urban
[07:56.000 --> 08:01.000]  vegetation you have at the same time for the heat islands it means with climate change
[08:01.000 --> 08:14.000]  it will get very hot in this areas. So but there's also the application directly in our
[08:14.000 --> 08:22.000]  real world technology that we're seeing and we found very amazing projects like a complete
[08:22.000 --> 08:30.000]  50 megawatt wind turbine that was being released open source so that means there's the blueprints
[08:30.000 --> 08:36.000]  how to build this wind turbine there's the simulation so all the software that actually is
[08:36.000 --> 08:43.000]  needed to design this wind turbines has been open sourced and in combination with multiple
[08:43.000 --> 08:53.000]  American and European organizations but you will also find interesting software libraries
[08:53.000 --> 08:59.000]  that help you to predict where do you place your photovoltaic and what is the perfect orientation
[08:59.000 --> 09:09.000]  of your photovoltaic in relation to the sun so that you have the most efficient of those solar
[09:09.000 --> 09:17.000]  panels but and there was also a talk about pipes yesterday we have now with open source
[09:17.000 --> 09:26.000]  it was we have now a complete simulation of the world energy grid and all those energy systems
[09:26.000 --> 09:35.000]  that are within this huge network of different power plants but also of the renewables and
[09:35.000 --> 09:45.000]  you see here for example the power grid of India how it was simulated with pipes so
[09:45.000 --> 09:54.000]  so what we actually did is a methodology that I from my analysis nobody did before an open
[09:54.000 --> 10:02.000]  source in general so the first thing we did is we compiled this list and it took me two years
[10:02.000 --> 10:09.000]  and I think more than 500 hours to investigate all those projects worldwide and we used multiple
[10:09.000 --> 10:16.000]  ways how to do it and you find it in the study in the report that we released some weeks ago
[10:16.000 --> 10:29.000]  it was a combination of researching on GitHub, GitLab, searching on different papers doing
[10:29.000 --> 10:36.000]  data mining but we also just looked on the stars of different people on GitHub so what are
[10:36.000 --> 10:42.000]  the people starring actually that are working in environmental sustainability so this was a
[10:42.000 --> 10:47.000]  huge effort but also the community helped us a lot so we had about more than 30 volunteers
[10:47.000 --> 10:54.000]  that contributed to the list and helped us to curate this list but and we really were focusing
[10:54.000 --> 11:03.000]  on projects that were aiming on environmental sustainability at their core. The core idea
[11:03.000 --> 11:10.000]  was to keep the list in a way that it is possible so we really try to list GitHub projects
[11:10.000 --> 11:19.000]  because GitLab and also GitLab projects are every kind of repository where people can
[11:19.000 --> 11:31.000]  participate into and so first step was creating the list and after that we created some scripts
[11:31.000 --> 11:38.000]  to automatically gather the metadata of all those projects and we created some targeted
[11:38.000 --> 11:46.000]  interviews with all those domains we investigated and the domains actually were derived within
[11:46.000 --> 11:53.000]  the study so this is not, this is our let's say perspective on this ecosystem and the
[11:53.000 --> 11:58.000]  namings that we use there and you see here in the middle that's actually the core project
[11:58.000 --> 12:04.000]  open sustainable technology and around us are the fields and within the fields are different
[12:04.000 --> 12:11.000]  topics that we put the projects into and after that we did the cooperative report and
[12:11.000 --> 12:18.000]  cross-validated actually the qualitative and quantitative insights that we derived from
[12:18.000 --> 12:32.000]  this investigation. So here's the overview of the data set and you can see we have
[12:32.000 --> 12:42.000]  in total 1,339 projects, this was quite some work and most of them were actually GitHub
[12:42.000 --> 12:50.000]  projects, some were GitLab projects and I must say actually the numbers here are all
[12:50.000 --> 12:58.000]  dealing with the GitHub projects. We could not integrate so far the GitLab API into the
[12:58.000 --> 13:04.000]  study, I hope to do this in further studies and also to integrate other platforms but
[13:04.000 --> 13:11.000]  to keep it simple the first study was really focusing on GitHub and actually on GitHub
[13:11.000 --> 13:20.000]  we also found most projects. You see here also that we listed the active project so
[13:20.000 --> 13:28.000]  within the study we tried only to list projects that had one closed issue in the last year
[13:28.000 --> 13:36.000]  that or one commit and that are documented and have a certain kind of quality so we did
[13:36.000 --> 13:43.000]  a small quality analysis of the project is there some documentation can people actually
[13:43.000 --> 13:48.000]  build the software by their own and use this in their own projects this was very important
[13:48.000 --> 13:55.000]  in the investigation and you can see here we derived a total number of stars and a total
[13:55.000 --> 14:00.000]  number of contributors but also other quantities to just show, give you an impression about
[14:00.000 --> 14:12.000]  how is the ecosystem in itself, how in this direction does it go. What you can also see
[14:12.000 --> 14:18.000]  is that we listed some inactive projects so actually this was the hard part of the study
[14:18.000 --> 14:26.000]  as you maybe know most open source projects die very soon so within the project actually
[14:26.000 --> 14:33.000]  192 projects actually got inactive and this was also my impression when we do this analysis
[14:33.000 --> 14:38.000]  most projects were really inactive. I have no clear numbers but I think it's more than
[14:38.000 --> 14:43.000]  90% of the projects we investigated were inactive they were just academia projects where somebody
[14:43.000 --> 14:49.000]  released a paper and then it's just over that's it it's like throwing a open source project
[14:49.000 --> 14:54.000]  to the wall and hope somebody maybe can use it but this is not where we want to focus on
[14:54.000 --> 14:59.000]  we really wanted to focus on a project that you can use and that you can reintegrate into
[14:59.000 --> 15:09.000]  newer projects. So the first thing that I realized when doing this investigation is
[15:09.000 --> 15:18.000]  there's quite a low popularity in environmental sustainability so even if it is a hot topic
[15:18.000 --> 15:23.000]  and you see it multiple times in the media and everybody is talking about it and yeah
[15:23.000 --> 15:36.000]  it sounds very trendy we could just find three projects that have more than 1,000 stars and
[15:36.000 --> 15:49.000]  in total on GitHub you find 38,000 projects that have more than 1,000 stars so it's just
[15:49.000 --> 15:54.000]  not and that's also what's my impression so most people are really not appreciating those
[15:54.000 --> 16:01.000]  climate models or for example those biodiversity models that have been created and that are
[16:01.000 --> 16:10.000]  actually being maintained by different organizations both white. So on the left hand side you see
[16:10.000 --> 16:19.000]  actually the color bar that is a community health index that we created it's called the
[16:19.000 --> 16:26.000]  development distribution score and we realized that actually the most important aspect of an
[16:26.000 --> 16:32.000]  open source project is how much is the knowledge and the development distributed within the
[16:32.000 --> 16:38.000]  project itself so we created just a very small number by the comment that is just we you take
[16:38.000 --> 16:45.000]  the number the comments of the strongest contributor and divide it by the total comments and you
[16:45.000 --> 16:52.000]  get an impression how much does this project actually depend on one developer or is there
[16:52.000 --> 16:59.000]  actually really community behind it and you see the AB Street that is the most popular
[16:59.000 --> 17:08.000]  project is really one-man show that's actually a project where somebody is trying to do a
[17:08.000 --> 17:19.000]  gamification of cities to improve our transport system to do it more sustainable. After that
[17:19.000 --> 17:24.000]  it's a very famous project and also it is a for-profit open source project that is called
[17:24.000 --> 17:34.000]  electricity map and after that we still see project like open farm or that or the scarf
[17:34.000 --> 17:39.000]  under that is actually a green software project so they are all dealing with different kinds
[17:39.000 --> 17:46.000]  of aspects how do we investigate and create knowledge about our environment and how can
[17:46.000 --> 17:54.000]  we improve our society so that we create less impact on our ecosystem let's say negative
[17:54.000 --> 18:02.000]  impact but you see also there some in popularity some climate models like or let's say earth
[18:02.000 --> 18:11.000]  science models and frameworks like Pangeo and it's or various models that are actually
[18:11.000 --> 18:20.000]  also very important like the WRF so on the right hand side you see the contributors so
[18:20.000 --> 18:25.000]  the total numbers of contributors that we could find within a project and electricity map
[18:25.000 --> 18:33.000]  there is again the let's say dominating project and electricity map is actually they combine
[18:33.000 --> 18:44.000]  the so-called carbon intensity of all different electric grids worldwide to give you an impression
[18:44.000 --> 18:50.000]  or give you a number about when I am consuming this amount of electricity at this point on
[18:50.000 --> 19:00.000]  the earth how much carbon do I actually release to the atmosphere by this power consumption
[19:00.000 --> 19:08.000]  and this was really let's say one of the very few larger communities that we could find
[19:08.000 --> 19:14.000]  where a lot of academic and for-profits and different organizations were coming together
[19:14.000 --> 19:22.000]  to create this the another project is actually the open food network where people try to create
[19:22.000 --> 19:33.000]  trustable knowledge and information about how what what is the information of food itself
[19:33.000 --> 19:38.000]  so that you can have trustworthy information and this is also a huge community project
[19:38.000 --> 19:45.000]  quite old and you can also see here just green project and very larger atmospheric project
[19:45.000 --> 19:50.000]  but in general you see the total numbers of contributors if you compare it with our open
[19:50.000 --> 20:03.000]  food projects are quite low it's not that big and for me personally actually the so this
[20:03.000 --> 20:09.000]  project itself has one thousand three hundred star so we are actually one of the most popular
[20:09.000 --> 20:13.000]  open source projects in the environmental sustainability and this is not a good sign
[20:13.000 --> 20:19.000]  but because I'm just we are just some random dudes it's we are we are actually we have
[20:19.000 --> 20:23.000]  a foundation we are just a community and people were coming together you know and that's what
[20:23.000 --> 20:29.000]  I wanted to show with this slide so within the open source community there's a lack for
[20:29.000 --> 20:41.000]  of popularity in environmental sustainability so if you look at the license and language
[20:41.000 --> 20:49.000]  of projects we see there's a lot of MIT license actually so it's actually easy to integrate
[20:49.000 --> 20:56.000]  those projects into for profit not that hard but there's also let's say a good portion
[20:56.000 --> 21:07.000]  of a free open source license like gpl 3.0 but what do we also could find is that with
[21:07.000 --> 21:13.000]  their data analysis we could not always really invest automatically investigate what is actually
[21:13.000 --> 21:20.000]  the license of the project so that's the custom license and this is always problematic so if
[21:20.000 --> 21:26.000]  you release open source code use a standard license that people know about they don't need a
[21:26.000 --> 21:33.000]  lawyer to actually see if they can actually reintegrate the software into maybe a commercial
[21:33.000 --> 21:39.000]  project because we want to when we need to go with such approaches into commercial projects
[21:39.000 --> 21:45.000]  if we don't commercialize environmental sustainability we will never achieve environmental
[21:45.000 --> 21:54.000]  sustainability on the right hand side you see the portion the proportions of the programming
[21:54.000 --> 22:01.000]  languages and here we can see that actually data analysis and data driven languages are
[22:01.000 --> 22:13.000]  really on the top I can imagine that you expect Python here but also R is here very prominent
[22:13.000 --> 22:21.000]  because in environmental sustainability it's a lot of statistics and processing a lot of data
[22:21.000 --> 22:30.000]  but we can also find here Fortron and C++ and those languages are used in the large scale
[22:30.000 --> 22:38.000]  simulation of our earth so if you want to really be efficient in your calculations that's the
[22:38.000 --> 22:44.000]  programming languages that you use and the high portion actually of Fortron shows us that many
[22:44.000 --> 22:52.000]  of those software projects are quite old this is really sometimes code that is 20 30 years old
[22:52.000 --> 22:59.000]  where our earth and our atmosphere biosphere hydrosphere has been simulated and this code has
[22:59.000 --> 23:09.000]  been reintegrated into new software projects yeah another JavaScript is here not that important
[23:09.000 --> 23:22.000]  so but that's what we also did is from the we actually derived from the projects all the
[23:22.000 --> 23:29.000]  organizations behind it so what is actually the namespaces of the projects and from the namespaces
[23:29.000 --> 23:35.000]  of the projects we had a little bit smaller list and this was actually hand labeled so we went
[23:35.000 --> 23:44.000]  through all those organizations and created actually the location but also the continent of
[23:44.000 --> 23:53.000]  those organizations and we see there's let's say Europe has quite a good standing here and also
[23:53.000 --> 23:58.000]  North America we found a lot of project that consider themselves organizations that consider
[23:58.000 --> 24:06.000]  themselves as global so but unfortunately I have to say actually from Asia and I did a lot of
[24:06.000 --> 24:14.000]  investigations I could not find a lot of open source projects that actually originate from the
[24:14.000 --> 24:18.000]  Asian continent whereas really an organization behind it it doesn't mean that there are not
[24:18.000 --> 24:23.000]  people involved into this just wanted to say that we could not find it and I did there a lot of
[24:23.000 --> 24:35.000]  research on this yes what we also did is we labeled the forms of organizations behind it so
[24:35.000 --> 24:42.000]  and it was surprising for me that actually the community based projects like we are the strongest
[24:42.000 --> 24:49.000]  so community we labeled when there was no legal structure behind it just some people coming
[24:49.000 --> 24:58.000]  together doing stuff this was considered community and after that academia has a very important
[24:58.000 --> 25:03.000]  role there because they did a lot of this research investigations in our environmental
[25:03.000 --> 25:10.000]  sustainability but also government agency are very important in this domain unfortunately
[25:10.000 --> 25:20.000]  I have to also to say for profits and that's a general trend are not that prominent we did
[25:20.000 --> 25:31.000]  I did a lot of investigations into the kind of do we find actually commercialized or for profit
[25:31.000 --> 25:36.000]  applications of this kind of software and I told you about electricity map that's maybe let's say
[25:36.000 --> 25:43.000]  the most one really successful project that we can find but besides that there is Microsoft is
[25:43.000 --> 25:48.000]  actually doing some good open source project in environmental sustainability like the planetary
[25:48.000 --> 25:56.000]  compute framework and so there's some other good developments but compared to the problem that
[25:56.000 --> 26:03.000]  we are dealing with this is a little bit disappointing so it cannot be that this is just
[26:03.000 --> 26:09.000]  been an academia community thing with some maybe some government agencies so there needs to be
[26:09.000 --> 26:14.000]  more for profit organizations and I think we see this also now with commercial open source
[26:14.000 --> 26:19.000]  software is becoming more and more important and there's really no billions of dollars going
[26:19.000 --> 26:26.000]  in this kind of development but we're missing there totally in this domain
[26:30.000 --> 26:41.000]  so you can see here this is all project that we mapped scattered over the project age you see
[26:41.000 --> 26:49.000]  here the topics and they are actually ordered we also created a size score so that we can order them
[26:49.000 --> 26:54.000]  you see the strongest was actually biosphere but this was also because we not separated
[26:54.000 --> 27:01.000]  actually biosphere so strongly if you would combine climate data processing and the climate models
[27:01.000 --> 27:10.000]  actually climate science I would say in general would be the most strongest area but also what
[27:10.000 --> 27:16.000]  was quite interesting is that actually hypersphere so the people dealing with our water resources
[27:16.000 --> 27:23.000]  they were also very strong we could find a lot of strong projects there it's all the same for soil
[27:23.000 --> 27:29.000]  it's all about air quality it's about agriculture and nutrition so really that what is really
[27:29.000 --> 27:34.000]  describing the state of our environment there we know a lot there's a lot of information about
[27:34.000 --> 27:41.000]  what is actually the state where we are going but if you go more into let's say the technology area
[27:41.000 --> 27:46.000]  if it's if you're talking about batteries if you're talking about hydrogen, bioenergy, carbon capture
[27:46.000 --> 27:56.000]  and removal there it was super hard to find good projects that we could list and also more
[27:56.000 --> 28:02.000]  surprisingly for me in the area of sustainable investment because sustainable investment is
[28:02.000 --> 28:07.000]  really something that a starter driven there's a lot of huge companies behind it that gather
[28:07.000 --> 28:14.000]  information that a lot of they have a lot of models and ways how to consider what is actually
[28:14.000 --> 28:22.000]  sustainable in our investment and there was actually two or three really strong projects
[28:22.000 --> 28:27.000]  but they were also at the beginning and compared to the problem we are facing this was also a little
[28:27.000 --> 28:38.000]  bit disappointing other areas like also carbon capture and carbon offsets were also not that prominent
[28:38.000 --> 28:46.000]  and I would say if you're more interested in the quantities that we created we created much more
[28:46.000 --> 28:51.000]  about the growth about the different kinds of communities and we created a lot of quantities
[28:51.000 --> 28:57.000]  and plots for you you will find us in the report so I would stop here to talk actually about
[28:57.000 --> 29:05.000]  let's say the quantitative analysis and go more into the into the suggestions into the recommendations
[29:05.000 --> 29:10.000]  that we derived from our analysis and from the interviews and from the people that we talked to
[29:10.000 --> 29:20.000]  and also I must say the discussions we had within the group lead to a lot of interesting aspects
[29:20.000 --> 29:31.000]  and recommendations so sustainable rating and investment is really getting a huge and important
[29:31.000 --> 29:38.000]  topic and it influence all of you because this is where actually your future is going into
[29:38.000 --> 29:47.000]  that means there's actually the United Nations principles for responsible investment this was
[29:47.000 --> 29:57.000]  created you see here 2006 and they created some really baseline for what and how do we actually do
[29:57.000 --> 30:05.000]  sustainable investment and you see here that actually in 2021 we are not talking about 100
[30:05.000 --> 30:14.000]  around 140 trillions of dollars that are invest invested based on those principles so there's
[30:14.000 --> 30:19.000]  a huge amount of money going in this direction that needs to be because we actually want to
[30:19.000 --> 30:33.000]  transform our economy to a let's say sustainable system so the problem is that how I was very
[30:33.000 --> 30:38.000]  interested in this domain and I did a lot of investigations I talked to people about this
[30:38.000 --> 30:45.000]  and we analyzed how this is done so how is actually sustainable rating and assessment
[30:45.000 --> 30:52.000]  being done and actually it is a big black box to summarize it you have some sustainability
[30:52.000 --> 30:59.000]  reports that are actually self reported by the companies and then are there's a lot of news
[30:59.000 --> 31:05.000]  about sustainability of companies and a lot of marketing that you know and this is actually
[31:05.000 --> 31:09.000]  being combined in so-called sustainability assessment and there are multiple companies
[31:09.000 --> 31:16.000]  that are doing this and there's really not so much known about really how this goes and you see
[31:16.000 --> 31:21.000]  here also how actually greenwashing works it's very easy you create very beautiful reports that
[31:21.000 --> 31:29.000]  really look very good nobody is really trace is really assessing if they are really based on
[31:29.000 --> 31:35.000]  science it's really that you do self reporting about your environmental sustainability and then
[31:35.000 --> 31:42.000]  this sustainability assessments also take into account news about those companies actually so how
[31:42.000 --> 31:47.000]  is there good news about the environmental impact of those companies or bad news and they also take
[31:47.000 --> 31:55.000]  financial news about those companies and this is all is being combined in sustainable sustainability
[31:55.000 --> 32:01.000]  assessment and then there comes a magic number and based on those numbers 140 trillion of
[32:01.000 --> 32:10.000]  dollars are invested somewhere okay that's crazy from my perspective and I dig a little bit into
[32:10.000 --> 32:19.000]  this and I realized oh there's a lot of scientists that realize that actually this is like throwing
[32:19.000 --> 32:26.000]  the dice what this rating companies are actually doing you see here on the bottom that's actually
[32:26.000 --> 32:32.000]  the value of the benchmark rating from one company called system analytics and then you see here the
[32:32.000 --> 32:40.000]  divergence of this ratings and if there would be an ideal word this would be one line okay they
[32:40.000 --> 32:48.000]  would all agree on environmental sustainability but and actually they do not agree so you just
[32:48.000 --> 32:53.000]  have to go to the right rating agency and you can make every company actually sustainable today
[32:53.000 --> 33:02.000]  very easily been done and you also have to see that actually the rating itself has some so you
[33:02.000 --> 33:09.000]  don't get normally the very bad rating or the very good rating right so in general it is you just
[33:09.000 --> 33:15.000]  have to throw the dice and also the offer saying addressing ESG rating that means environmental
[33:15.000 --> 33:19.000]  governments because this is actually a little bit a combination of different kinds of ratings they do
[33:19.000 --> 33:28.000]  environmental but also a social sustainability rating and governments rating at diverges requires
[33:28.000 --> 33:34.000]  addressing ESG rating divergence requires one to understand how the data that underpins ESG
[33:34.000 --> 33:43.000]  ratings are generated so nobody knows how this data is generated and I think you know already the
[33:43.000 --> 33:52.000]  solution so when nobody knows our data is generated and we have a big black box yeah then
[33:52.000 --> 33:58.000]  you normally take open source but this is not happening because people are also doing the same
[33:58.000 --> 34:04.000]  thing in the so-called carbon offset so if your company is actually creating a lot of emissions
[34:04.000 --> 34:09.000]  and you want to be carbon neutral and you see this labeled all over the place now everybody's
[34:09.000 --> 34:17.000]  carbon neutral and it's great that they all actually want to be carbon neutral but there's
[34:17.000 --> 34:24.000]  actually a study being done by the Guardian and they investigated how much of those carbon that is
[34:24.000 --> 34:30.000]  actually released in the atmosphere is actually being stored back into our soil and into our
[34:30.000 --> 34:35.000]  ground and you see this on the right hand side there's actually the carbon cycle that is driving
[34:35.000 --> 34:42.000]  the CO2 into the atmosphere is actually a natural let's say there's a natural cycle that actually
[34:42.000 --> 34:54.000]  is going on and this cycle was created by actually it took millions of years to stabilize our
[34:54.000 --> 34:59.000]  atmosphere by this carbon cycle so that we have a low portion of carbon within our atmosphere
[34:59.000 --> 35:05.000]  and this way a stable weather and stable climate and the predictable earth systems that we all
[35:05.000 --> 35:13.000]  can depend on but yeah the Guardian found out that the real emissions that are actually stored
[35:13.000 --> 35:22.000]  in the ground are much lower than actually what they are claiming for so actually they say that
[35:22.000 --> 35:33.000]  over 90% of those carbon offsets people are claiming for actually claims that's what the
[35:33.000 --> 35:37.000]  Guardian is saying here but you see another problem Guardian is the Guardian is just showing
[35:37.000 --> 35:43.000]  your plot and that's actually the core problem of all of this they just at the end show your
[35:43.000 --> 35:50.000]  plot and then you believe my plot I did great science behind it just trust my science and
[35:50.000 --> 35:57.000]  that's let's say the core problem I want to investigate with you a little bit more so
[35:57.000 --> 36:05.000]  because we're talking about here about safety critical decision making the we know exactly
[36:05.000 --> 36:10.000]  and I showed you this also with the topics and with the numbers of people that are investigating
[36:10.000 --> 36:16.000]  we know very good with open science that is traceable for all of you you can go into this
[36:16.000 --> 36:23.000]  climate models and they're very good examples for in in in simple Python you can but you
[36:23.000 --> 36:33.000]  will also find larger open climate science projects that you can just run on your machine
[36:33.000 --> 36:40.000]  and they all come to the same result that we are actually on a very bad path on with
[36:40.000 --> 36:47.000]  colliding with the earth actually and this is very clear to us so there's no doubt so
[36:47.000 --> 36:54.000]  if you if you don't trust into climate science just go into repositories create an issue or
[36:54.000 --> 37:01.000]  deal with the software that you find there so opens the problem that we are facing now
[37:01.000 --> 37:08.000]  is that the decision making the assessment of sustainable and sustainable investment there
[37:08.000 --> 37:14.000]  we have no clue what we actually do and there's no open source software and not much open
[37:14.000 --> 37:25.000]  source software that we could really find so but there's help and especially in the energy
[37:25.000 --> 37:32.000]  sector we found a lot of good projects that that that really understand how to get rid
[37:32.000 --> 37:39.000]  of those problem of traceability because when somebody just is showing you a plot I think
[37:39.000 --> 37:46.000]  hydrogen will be the future of energy of the energy sector and there comes somebody else
[37:46.000 --> 37:51.000]  and shows you another plot this cannot really come you cannot really create a discussion
[37:51.000 --> 37:57.000]  about this and this is really everybody needs to be at least able to discuss about this right
[37:57.000 --> 38:05.000]  now and what they did in in actually in the energy modeling so the this pipes up project
[38:05.000 --> 38:10.000]  actually I showed you at the beginning they release those models and in this way you can
[38:10.000 --> 38:19.000]  create proof and show hey I calculated this if you don't agree with me show me your model
[38:19.000 --> 38:27.000]  and if you don't show me your model I don't trust you that's that's how it goes and this
[38:27.000 --> 38:38.000]  is really let's say a very important idea that I found here it's that you you need to
[38:38.000 --> 38:46.000]  understand that you can actually prove how much conclusion is actually being open and
[38:46.000 --> 38:51.000]  traceable you just have to see how much open data is there and is there uncertainties
[38:51.000 --> 38:57.000]  actually also in the data that is comes with the data is there open source models are there
[38:57.000 --> 39:03.000]  is there actually an open execution possible of those models do you find open results of
[39:03.000 --> 39:08.000]  those execution so is the artifacts actually create by its simulation is this actually
[39:08.000 --> 39:13.000]  also open source is there a way to participate can you go into the GitHub repository can you
[39:13.000 --> 39:19.000]  go into a conference and ask the question how you had has he done this and is this conclusion
[39:19.000 --> 39:25.000]  with uncertainties at the end so is there a plot that shows what is actually my unknown
[39:25.000 --> 39:31.000]  my uncertainty what or is it just a single number and if it's just a single number it's
[39:31.000 --> 39:39.000]  a problem and we can propagate the uncertainty through such a thing and that's very important
[39:39.000 --> 39:49.000]  for traceability then what we found out that actually openness itself and that's let's say
[39:49.000 --> 39:56.000]  one of the core findings of the whole study is the is the indicator for sustainability
[39:56.000 --> 40:02.000]  itself so if you want to know and it's very important to understand we sometimes don't
[40:02.000 --> 40:10.000]  know today what is actually sustainable but if there comes somebody and says I personally
[40:10.000 --> 40:15.000]  don't know if I'm really sustainable but I show you how I calculated why I think I am
[40:15.000 --> 40:21.000]  sustainable this show sustainable intention and this is what we need today because then
[40:21.000 --> 40:31.000]  you can go into and discussion you can go with people into some maintenance of code
[40:31.000 --> 40:36.000]  you can improve your own code and that's why actually openness itself is the core is
[40:36.000 --> 40:43.000]  really the key indicator for sustainability itself you see also despite a picture this
[40:43.000 --> 40:50.000]  was shows the standing on the shores of giants but from my perspective really describes very
[40:50.000 --> 40:55.000]  good open source open source or let's say the technology that we built that is this
[40:55.000 --> 41:02.000]  blind giant and we are standing on the shoulder but we can see farther by the giant but you
[41:02.000 --> 41:08.000]  cannot but the giant itself is blind but we can see much much farther and we can and
[41:08.000 --> 41:13.000]  the giants get bigger and bigger and we can see and predict the future in a much better
[41:13.000 --> 41:17.000]  way and that's why open source and actually the math and the models that we create there
[41:17.000 --> 41:24.000]  are so important to predict the future so one interesting aspect that I found to get
[41:24.000 --> 41:30.000]  rid of this problem is actually spatial finance it's the idea that we can use those
[41:30.000 --> 41:35.000]  whole earth observation data that we have to track down companies about environmental
[41:35.000 --> 41:44.000]  sustainability so all those software in earth observation all those open source software
[41:44.000 --> 41:51.000]  there is a huge amount of open source software and earth observation can be used by rating
[41:51.000 --> 41:57.000]  agencies and all of us to actually track down companies on the environmental impact that's
[41:57.000 --> 42:04.000]  not that hard and there's actually an Oxford group that started the spatial finance and
[42:04.000 --> 42:11.000]  what they started with is they created a map of the whole fossil fuel industry where you
[42:11.000 --> 42:17.000]  know the whole supply chain so where are they located where is the fuel supply chain that
[42:17.000 --> 42:22.000]  helps you to on the one side simulate it but also to check satellite data and track down
[42:22.000 --> 42:26.000]  okay this company has a claim that they're actually sustainable let's have a look and
[42:26.000 --> 42:35.000]  you have suddenly a sensor that is measuring the sustainability and that is needed so here
[42:35.000 --> 42:46.000]  is my let's say that's how I would actually improve the rating of environmental sustainability
[42:46.000 --> 42:53.000]  and how we can actually get rid of those problem is to actually integrate earth observation
[42:53.000 --> 43:00.000]  into this processing chain and very importantly where the assessment happens we need open source
[43:00.000 --> 43:07.000]  code and that's actually just one from the Leung's Foundation one project that is starting
[43:07.000 --> 43:12.000]  to do this but this is really everything very early stage and then suddenly we can maybe
[43:12.000 --> 43:18.000]  in the future get sustainability ratings about companies that have some uncertainties but
[43:18.000 --> 43:24.000]  it would be also very great if companies itself would use open source code to do environmental
[43:24.000 --> 43:29.000]  sustainability rating within the company so for me personally I would like if ExoMobil
[43:29.000 --> 43:34.000]  would say we would go open science now or every company that goes open science about their
[43:34.000 --> 43:39.000]  own environmental sustainability would have such a massive impact on the industry because
[43:39.000 --> 43:46.000]  finally you have the information and data to really measure sustainability within the
[43:46.000 --> 43:54.000]  company because so far this is all the black box for us so yeah I would like to say that
[43:54.000 --> 44:00.000]  open source is from my perspective the most underestimated climate strategy or climate action
[44:00.000 --> 44:10.000]  it is there's so much opportunity there that we created a whole list of recommendations
[44:10.000 --> 44:17.000]  how open source can have a significant impact so actually we were thinking about what could
[44:17.000 --> 44:22.000]  be follow-up projects and we had so many ideas about follow-up projects that we just were
[44:22.000 --> 44:29.000]  starting to list them all down as recommendations I think there are more than 50 now and I just
[44:29.000 --> 44:35.000]  thought you two of them in combination actually so and yeah my final words would actually
[44:35.000 --> 44:42.000]  be and that's I think we need to understand we need to build an operating system for environmental
[44:42.000 --> 44:51.000]  sustainability it is for this for things like carbon offsets or the rating of the environmental
[44:51.000 --> 44:57.000]  impact of companies you need to connect all the spheres all those open source projects
[44:57.000 --> 45:05.000]  together to get one to bet get a better conclusion about what is the state of our environment
[45:05.000 --> 45:12.000]  and that's why and that's I hope and that's what we hope to start and hope to build in the
[45:12.000 --> 45:38.000]  future is really such kind of an operating system yeah thank you so much
[45:38.000 --> 45:48.000]  well thank you Tobias for your brain exploding presentation we have you can type your questions
[45:48.000 --> 45:56.000]  in matrix or you can just raise your hand and I will bring you the microphone oh gosh okay
[45:56.000 --> 46:16.000]  I lost you in the crowd again again please Tobias thank you so much for very inspiring
[46:16.000 --> 46:26.000]  presentation can you hear me and I absolutely agree that open source is very much underestimated
[46:26.000 --> 46:36.000]  as source of sustainability and a question can you suggest any ways how can developers
[46:36.000 --> 46:45.000]  collaborate better so for example I know that there are some gaps in climate modeling in
[46:45.000 --> 46:51.000]  bridging between climate modeling and energy modeling and it would be very helpful if energy
[46:51.000 --> 46:58.000]  modelers would know better about climate modelers and vice versa do you have any ideas can you
[46:58.000 --> 47:07.000]  suggest something to improve this situation can you hear me yeah I think what is really
[47:07.000 --> 47:13.000]  missing is the community behind it that is cross over the topics actually conference you
[47:13.000 --> 47:18.000]  meet somewhere where these people meet and talk to each other that's what is missing and
[47:18.000 --> 47:24.000]  we realized and that's why we actually would and also the topic of environmental sustainability
[47:24.000 --> 47:30.000]  is always a side topic in open source conferences somewhere but we really need a conference
[47:30.000 --> 47:36.000]  actually there that's the point where this this connection between the software projects is
[47:36.000 --> 47:42.000]  starting at the beer in the evening maybe where these people are connecting they are not really
[47:42.000 --> 47:52.000]  connected with each other that's why this list is so successful they could not find each other
[47:52.000 --> 48:01.000]  Tobias you mentioned that's a very low for profit organizations currently yes but this if they
[48:01.000 --> 48:09.000]  increase we have also the lobbyism problem again because they are much more trained to get
[48:09.000 --> 48:17.000]  money for their already existing projects do you see some chance that new open source projects
[48:17.000 --> 48:29.000]  in this group will also get paid contributors this is also an opportunity yes it's not I'm not
[48:29.000 --> 48:33.000]  the biggest fan of commercial open source software I see there's some opportunities there are
[48:33.000 --> 48:40.000]  good ways bad ways it depends on topic and the climate model should maybe not be in for profit
[48:40.000 --> 48:46.000]  but if it is maybe about the rating yeah there should be a company behind so something it really
[48:46.000 --> 48:53.000]  depends on the kind of project we are talking about so for example if you if you want to create
[48:53.000 --> 48:59.000]  a wind turbine right a wind turbine that is open source completely you need a company that
[48:59.000 --> 49:06.000]  maybe gives you 40 years of maintenance and it needs to be sure that this is maintenance and
[49:06.000 --> 49:11.000]  miss maybe this is better with a for profit but depending on the project maybe on with a
[49:11.000 --> 49:16.000]  non-profit it's not an easy question what is the right organization for an open source project
[49:16.000 --> 49:30.000]  it's very complicated