[00:00.000 --> 00:10.300]  Hello everyone at the energy left room at FOSDEM.
[00:10.300 --> 00:14.680]  My name is Felix Riemann and I'm happy to be here and present you some insights from
[00:14.680 --> 00:19.640]  projects at Einstein Center for the Future and Theo Berlin, both located in the heart
[00:19.640 --> 00:21.200]  of Berlin.
[00:21.200 --> 00:27.480]  And today I'm focusing on some technologies and some obstacles we find dealing with energy
[00:27.480 --> 00:33.360]  technologies and related open source software in the German research landscape.
[00:33.360 --> 00:38.280]  To give you a slight introduction, I'm going today to talk about why do we need open source
[00:38.280 --> 00:44.320]  software for buildings and energy, how do buildings impact climate, how can open source
[00:44.320 --> 00:48.640]  reduce that impact and what is the current state of open source.
[00:48.640 --> 00:52.400]  I'm going to give you a short brief review of funded project.
[00:52.400 --> 00:58.760]  I'm going to give you some major obstacles we identified with the usage of open source
[00:58.760 --> 01:04.960]  and how researchers are applying it currently and then last but not least, I'm going to
[01:04.960 --> 01:08.560]  give you an outlook where we will be in five years.
[01:08.560 --> 01:13.720]  But let's start with a short introduction of our job and our goal and why we think FOSDEM
[01:13.720 --> 01:16.680]  is the ideal place to be here.
[01:16.680 --> 01:22.560]  So we are doing, we are the accompanying research.
[01:22.560 --> 01:28.920]  So what we are doing is we are supporting more than 300 research projects and survive
[01:28.920 --> 01:36.920]  them merely through different means and our focus on different aspects of digitalization.
[01:36.920 --> 01:43.800]  So for example, we look at data governance, we look at which tools they apply, for what
[01:43.800 --> 01:49.600]  reasons do they apply these tools and we try to support and connect projects and this
[01:49.600 --> 01:54.040]  is especially why we're happy to be here at FOSDEM because we are coming from an energy
[01:54.040 --> 01:59.200]  perspective and we would like to get to know more people in open source community and get
[01:59.200 --> 02:04.400]  feedback from you and foster the exchange between the national research community and
[02:04.400 --> 02:10.280]  the international open source community to learn and foster the exchange.
[02:10.280 --> 02:17.960]  Our goal is supporting standardization and integration of software and standards so that
[02:17.960 --> 02:26.800]  other people can apply the solutions, we help our researchers that we support and reuse.
[02:26.800 --> 02:31.840]  And why do we think residential or buildings in general and neighborhoods are especially
[02:31.840 --> 02:34.640]  relevant for that case?
[02:34.640 --> 02:43.840]  So if we look in Germany around 35% of the end energy usage are related to buildings
[02:43.840 --> 02:51.000]  and around two thirds of that are related to residential buildings and the majority of
[02:51.000 --> 02:59.920]  that is for heating and as you can see here on the left there's some empirical data from
[02:59.920 --> 03:08.080]  buildings, energy usage, a lot of that energy usage is actually used in old buildings.
[03:08.080 --> 03:14.120]  And what we can see from that, that older buildings use more energy than new buildings
[03:14.120 --> 03:21.640]  because of variety of reasons e.g. different installation standards but also we can see
[03:21.640 --> 03:27.760]  that we need to focus on a specialty of buildings especially older buildings and keep in mind
[03:27.760 --> 03:36.720]  that with the long lifespan of buildings around 30 plus years as you can see here that tomorrow
[03:36.720 --> 03:42.320]  is building a built today so if we want to be climate neutral by 2045 we need to build
[03:42.320 --> 03:45.760]  climate neutral buildings now.
[03:45.760 --> 03:53.040]  And open source technology can help with that for variety of reasons especially in three
[03:53.040 --> 03:54.040]  strategies.
[03:54.040 --> 04:00.360]  They can help reduce the demand through installation of better insulation so if you for example
[04:00.360 --> 04:06.200]  have open planning tools you can choose the right approach and where to insulate, what
[04:06.200 --> 04:12.960]  to insulate and what is cost effective and can also replace technology you can find for
[04:12.960 --> 04:18.520]  example the perfect heat pump for your place which can replace the boiler you are using
[04:18.520 --> 04:24.240]  or you can also have better control strategies which is mostly applied in non-residential
[04:24.240 --> 04:30.000]  building so if you for example have a demand or crit orientated supplying approach that
[04:30.000 --> 04:37.200]  also reduces the total energy demand or makes it more crit friendly.
[04:37.200 --> 04:44.240]  So digital technologies are essential for climate friendly buildings for the three reasons
[04:44.240 --> 04:51.920]  and we need software to help to plan and run these buildings and our impression or our
[04:51.920 --> 04:59.280]  goal is to make this software open source because it can foster the transformation, it can make
[04:59.280 --> 05:07.800]  it cheaper, it can make it more transparent and also faster and how can they do that?
[05:07.800 --> 05:13.520]  For example I have five different light phases in a building this is for example the plan
[05:13.520 --> 05:20.000]  which we start normally and for example here we can apply open computer edit design so
[05:20.000 --> 05:26.640]  if we have better orientation in a building you also need less energy demand because the
[05:26.640 --> 05:29.480]  sun is actually helping to heat our work.
[05:29.480 --> 05:35.160]  If we are dealing with a build environment and we have a variety of tasks for example
[05:35.160 --> 05:41.320]  as build classification so we know which boiler is actually built in in the building and not
[05:41.320 --> 05:50.120]  only plan to build in which is huge hassle or huge hassle in a lot of actually building
[05:50.120 --> 05:56.200]  environments we're going to go in the usage phase monitoring can be used to deploy building
[05:56.200 --> 06:00.960]  control strategies we can for example see oh the heater is running while we actually
[06:00.960 --> 06:08.880]  also have an open window which is suboptimal so we can identify faults and if we look at
[06:08.880 --> 06:14.720]  the renovating phase of a building we can actually choose the right fit between technology
[06:14.720 --> 06:21.520]  and insulation and that's but not least if we go to more life cycle oriented approach.
[06:21.520 --> 06:25.760]  Material databases can improve the recycling quarter for example we need to identify which
[06:25.760 --> 06:33.280]  material is built where and where and how is it used or how can we reuse it and how is
[06:33.280 --> 06:39.760]  the German Federal Ministry of Economics and Fair helping to shape that so as you can
[06:39.760 --> 06:45.000]  see on the left there's roughly spend more than 100 million each year on research projects
[06:45.000 --> 06:55.000]  so that's quite a lot and just a short takeaway from this slide in the last few years it's
[06:55.000 --> 07:06.680]  been shifting from buildings to neighborhoods and to heating this networks and we say and
[07:06.680 --> 07:15.080]  we also can think if you look at general information that integration and linkage of different technologies
[07:15.080 --> 07:20.320]  is becoming more and more important and for example hence the funding for neighborhoods
[07:20.320 --> 07:28.680]  has been growing instead of isolated topics and what are these topics that are being developed
[07:28.680 --> 07:38.520]  so in 2021 we did the survey and we survived 179 projects and out of these projects around
[07:38.520 --> 07:45.160]  128 said they were developing or using some kind of digital applications.
[07:45.160 --> 07:51.120]  If you're wondering why there's a huge gap to be honest we are also wondering but we
[07:51.120 --> 07:59.800]  also asked the projects to exclude the answers if they are for example using Excel as a software
[07:59.800 --> 08:05.440]  data developing and we see that actually quite often because if you're focused for an example
[08:05.440 --> 08:13.440]  on energy planners which are mostly self-employed very very tiny bureaus they actually use quite
[08:13.560 --> 08:17.840]  a lot of Excel so that's actually helping them but it's not the kind of software we
[08:17.840 --> 08:24.640]  are looking at so the software is excluded here and the kind of topics you see on the
[08:24.640 --> 08:29.360]  left side is mostly focused on simulation, operation optimization, monitoring, energy
[08:29.360 --> 08:36.040]  management, learning tools and what you also can see that none of the project is actually
[08:36.040 --> 08:43.200]  focused on social aspect so there's a huge gap in that on how we can include social aspects
[08:43.200 --> 08:48.320]  in digital aspects together in buildings and neighborhoods so if you're looking for
[08:48.320 --> 08:58.320]  a research project there it is and what did the project think about open source?
[08:58.320 --> 09:05.040]  A previous project by some colleagues of us did a survey in 2018 and they found that only
[09:05.040 --> 09:12.400]  minority of software or roughly 3% of Soviet projects planned for a full open source release
[09:12.400 --> 09:20.200]  of developed tools and software so that's not a lot isn't it and most of the software
[09:20.200 --> 09:26.040]  uses at least one proprietary tool so keep in mind that if they want to build software
[09:26.040 --> 09:31.840]  they use an average 4-5 other digital tools we call it a tool chain and I will talk later
[09:31.840 --> 09:40.960]  about this a bit more and the majority of these tools around 70% are not open source
[09:40.960 --> 09:46.560]  so there's not a lot of open source being released and there's also not a lot of open
[09:46.560 --> 09:56.080]  source being used what might be some reasons for that so there's a paper by some colleagues
[09:56.080 --> 10:02.440]  Stefan Feninger and all the papers from 2017 and they found some variety of reasons they
[10:02.440 --> 10:09.960]  focused on data and open source software and some reasons they find is that ag ethical
[10:09.960 --> 10:16.240]  and security concerns so there's sensitive and personal information might be included
[10:16.240 --> 10:22.960]  and through a variety of reasons everyone is afraid they might overshare personal and
[10:22.960 --> 10:28.960]  sensitive information unwanted exposure so if your stuff is public everyone can find
[10:28.960 --> 10:34.800]  the mistakes you make personally I think it's important to have stuff public because no
[10:34.800 --> 10:42.960]  one can find the mistakes and someone is referring to you and then they are repeating your mistakes
[10:42.960 --> 10:50.960]  so it's important to have stuff open so mistakes can be identified but that might also be a
[10:50.960 --> 10:57.120]  reason for some people not to have it because some of them might identify mistakes then we have to
[10:57.120 --> 11:02.400]  protection of intellectual property and I will talk about this a little bit more in the future but
[11:02.400 --> 11:07.840]  expertise is a business model or can be a business model and then we have institutional
[11:07.840 --> 11:15.920]  personal inertia so long-running practice are very hard to overcome especially for huge organizations
[11:15.920 --> 11:24.240]  that apply a different standard in a different methodology and let's continue with obstacles
[11:24.240 --> 11:29.600]  we find in our research landscape and especially related to buildings as this paper was more focused
[11:29.600 --> 11:38.480]  on energy systems engineering so I categorized in three categories starting with technical
[11:38.480 --> 11:46.480]  obstacles I will continue with cultural obstacles and finish with financial obstacles so especially
[11:46.480 --> 11:53.360]  in buildings and related technology we have a huge heterogeneity in data as you saw on a previous
[11:53.360 --> 11:59.680]  slide we have quite an age gap with a variety of energy demand variety of energy related technology
[12:00.240 --> 12:08.080]  and this also makes the software people deploy or develop for it quite heterogeneous because
[12:08.080 --> 12:14.640]  you have to for example identify different data points you have to example different ICT and so on
[12:14.640 --> 12:20.720]  so on the one hand some may think it's not worth publishing software anyway on the other hand some
[12:20.720 --> 12:26.480]  software may not be applicable at all because it's focused on a very special aspect so people think
[12:26.480 --> 12:32.240]  oh why make it open source but we think hey there might be at least a second person and it's public
[12:32.240 --> 12:42.800]  money then the two chains so if one part of the two chain is not documented at all or not open
[12:42.800 --> 12:49.520]  source or open science it cannot be reused and we think we need more modular and more and better
[12:49.520 --> 12:56.880]  documented two chains so the individual components need to be documented and well understood and then
[12:56.880 --> 13:02.400]  integrated in a chain instead of focusing on a complete chain and paper people should focus
[13:02.400 --> 13:09.280]  on the components and then we have missing open software which is especially relevant to cat
[13:09.280 --> 13:17.920]  and solvers and in some areas there's just like a huge gap and this leads to technical obstacles
[13:17.920 --> 13:27.520]  in making the complete system open source and as we can see on the right basic technical
[13:29.200 --> 13:36.080]  prerequisites is that the interfaces fit and we think that is relevant for all of these technical
[13:36.080 --> 13:42.720]  obstacles so we need to have fitting interfaces for data for the tool chains and last but not least
[13:42.720 --> 13:48.640]  for the missing open source software so we can have a whole ecosphere of software
[13:50.240 --> 13:56.480]  let's continue with the cultural obstacles we have a variety of cultural obstacles identified
[13:57.680 --> 14:03.600]  and one might be surprising but we actually have a lack of development skill in our understanding
[14:03.600 --> 14:07.840]  so software development is especially in mechanical engineering education
[14:08.560 --> 14:16.640]  and they're represented at least in Germany so that's our impression and also this leads to
[14:17.520 --> 14:20.800]  lack of common criteria for software quality
[14:21.360 --> 14:22.800]  so
[14:35.520 --> 14:39.840]  additionally in research software is often seen as a tool rather than an output
[14:40.560 --> 14:45.760]  so the researchers focus on publishing papers instead of publishing software
[14:46.480 --> 14:51.840]  and publishing the software is often overlooked or doesn't happen at all
[14:52.800 --> 14:59.040]  and last but not least I think this is more has more to do with the usage phase of software
[14:59.760 --> 15:06.640]  we did a survey in 2021 as previously said and out of the surveyed projects
[15:07.200 --> 15:14.080]  and only a minority said they are testing the software with their users so
[15:15.680 --> 15:22.320]  this in our opinion reduces the applicability and even if the software is open source no one
[15:22.320 --> 15:28.240]  is using it because it's not tested evaluated and the users don't understand it so we need a bigger
[15:28.240 --> 15:36.080]  um responsibility or bigger focus on using testing so the people that are supposed to
[15:36.080 --> 15:46.560]  use the software in the end actually using it last but not least we have the financial
[15:46.560 --> 15:53.840]  obstacles and especially the first one is relevant in our case so we quite often have
[15:53.840 --> 16:01.280]  enter funding so when a phd is done or project is finished and then you have to write your report
[16:01.280 --> 16:09.920]  but the report is just text and with that often uh no one any focus anymore on developing or
[16:09.920 --> 16:16.400]  publishing the software and we need to find a financial structure or funding structure
[16:17.040 --> 16:23.840]  that focus on also publishing all outputs in a well-documented and well-understood way
[16:24.480 --> 16:31.120]  then we have some business interest research is also a business having tools especially
[16:31.120 --> 16:38.560]  only for you can help you get new funding and can help you stand out and quite often we now
[16:38.560 --> 16:43.520]  see commercial alternatives with especially project partners wanting to have a long-term
[16:43.520 --> 16:52.880]  service agreement so they choose rather like um commercial alternative and then uh we have
[16:53.600 --> 16:59.520]  as they can provide a long-term service currently the maintenance the and the use of the
[16:59.520 --> 17:06.720]  software development of the software often ends with the project um participant of a workshop
[17:06.720 --> 17:14.080]  to us that we think it's important to find value behind project to have a community around open
[17:14.080 --> 17:20.400]  source deployed and this is why we need to focus on a value added to practice beyond your research
[17:20.400 --> 17:29.360]  project let me talk about the toolchain example as previous said a little bit more so quite often
[17:30.000 --> 17:35.920]  we have two people looking to build the software for the same solution so example you see user
[17:36.160 --> 17:42.560]  a up here and he wants to build um software for demonstration and user b or she wants to
[17:42.560 --> 17:49.520]  reuse that software but she's missing the license so she cannot reuse the tool or the software
[17:49.520 --> 17:59.440]  completely and she has to rebuild it so we have a lot of lost um power and we have a lot uh
[17:59.440 --> 18:05.200]  inefficiencies due to that and if you keep in mind that on average four or five tools are used to
[18:05.200 --> 18:11.200]  build software in your research project this happens actually quite often so we think partial
[18:11.200 --> 18:19.760]  open source is not enough we need modularized and well understood open source and only this
[18:19.760 --> 18:26.800]  can foster an open source culture as the software is only open source when all of the parts so to
[18:26.800 --> 18:31.840]  speak the complete toolchain are understood well documented and open source themselves
[18:34.480 --> 18:41.680]  this leads me to the recommendations i only brought six but we can discuss this later so
[18:42.800 --> 18:48.320]  as already said with the two chains i think modularized development is important
[18:49.760 --> 18:55.600]  so if you have each of the building stones of your overall software well understood
[18:55.680 --> 19:01.280]  documented and open source this enables maintenance and reusability of the whole software
[19:02.400 --> 19:10.160]  then we have user focused or more user focused software development actually the um software
[19:10.160 --> 19:17.040]  can be used and redeveloped and maintained after the project and we also need to focus on
[19:17.040 --> 19:23.760]  standardized interface interfaces especially with building technology and buildings as everyone
[19:23.760 --> 19:30.560]  is using some kind of object orientation but not the same so we cannot apply the same interfaces
[19:30.560 --> 19:38.880]  and so on and so on and so on and then we think it's important to publish this data scroll so
[19:38.880 --> 19:46.240]  actually we already have a lot of software and documentation and related papers but there's no
[19:46.240 --> 19:52.400]  general web page currently that publishes everything and collects everything that is being
[19:53.040 --> 20:01.600]  funded by public money so that's also an important step then also focusing on essentials crucial
[20:01.600 --> 20:09.520]  software e.g monitoring where which every project uses and every project or every building can use
[20:09.520 --> 20:17.760]  to save energy should be focused on and last but not least adapted funding because for maintaining
[20:17.760 --> 20:25.840]  software we often need ongoing funding structures let's come let me come to the last part of the
[20:25.840 --> 20:34.400]  presentation so you heard a lot about obstacles and i gave you a few brief introduction on
[20:34.400 --> 20:40.480]  what are missing stepping stones so far in the past but what's the current state of open source
[20:40.480 --> 20:46.640]  so on the left you see an argos which is a database where publicly funded projects
[20:47.680 --> 20:54.720]  are listed and if you look up open source you find around 446 projects at least in the beginning of
[20:54.720 --> 21:02.160]  January and the positive thing is there has been a sharp increase in recent years only 16 projects
[21:02.160 --> 21:12.000]  of these 446 projects have been started before 2010 but is quite often unclear what is considered
[21:12.000 --> 21:17.040]  open source and especially keep in mind the two chains is only the hardest open source
[21:17.680 --> 21:25.840]  everything can still be not reusable in open source way but i think there's been a sharp increase
[21:25.840 --> 21:32.800]  and we have a variety of open source projects listed and where we'll be in the next years so
[21:34.240 --> 21:40.720]  if we think about the future like for decarbonization and digitization of the energy system
[21:42.800 --> 21:48.800]  we need quite some building stones e.g infrastructure high quality data involvement of
[21:48.800 --> 21:58.640]  people as already said and if we did a server again in November 2020 and we asked around 250
[21:58.640 --> 22:04.640]  researchers if there needs a promotion of open standard tools like energy plus and corresponding
[22:04.640 --> 22:13.680]  libraries and the majority of them agreed so with that outlook that everywhere that we think a lot
[22:13.680 --> 22:19.680]  of researchers and even the publicly funded structure is shifting to more open source approach
[22:20.720 --> 22:27.360]  and is considering the need for integration of these different opening open source approaches
[22:27.360 --> 22:34.720]  let me finish the presentation i'm happy to ask any questions here in the chat or per email and
[22:34.720 --> 22:41.120]  thank you very much for your attention and have a great day at FOSDEM